US20060254411A1 - Method and system for music recommendation - Google Patents
Method and system for music recommendation Download PDFInfo
- Publication number
- US20060254411A1 US20060254411A1 US11/492,355 US49235506A US2006254411A1 US 20060254411 A1 US20060254411 A1 US 20060254411A1 US 49235506 A US49235506 A US 49235506A US 2006254411 A1 US2006254411 A1 US 2006254411A1
- Authority
- US
- United States
- Prior art keywords
- song
- music
- user
- songs
- file
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Abandoned
Links
Images
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10H—ELECTROPHONIC MUSICAL INSTRUMENTS; INSTRUMENTS IN WHICH THE TONES ARE GENERATED BY ELECTROMECHANICAL MEANS OR ELECTRONIC GENERATORS, OR IN WHICH THE TONES ARE SYNTHESISED FROM A DATA STORE
- G10H1/00—Details of electrophonic musical instruments
- G10H1/0033—Recording/reproducing or transmission of music for electrophonic musical instruments
- G10H1/0041—Recording/reproducing or transmission of music for electrophonic musical instruments in coded form
- G10H1/0058—Transmission between separate instruments or between individual components of a musical system
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/60—Information retrieval; Database structures therefor; File system structures therefor of audio data
- G06F16/63—Querying
- G06F16/632—Query formulation
- G06F16/634—Query by example, e.g. query by humming
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/60—Information retrieval; Database structures therefor; File system structures therefor of audio data
- G06F16/63—Querying
- G06F16/635—Filtering based on additional data, e.g. user or group profiles
- G06F16/637—Administration of user profiles, e.g. generation, initialization, adaptation or distribution
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/60—Information retrieval; Database structures therefor; File system structures therefor of audio data
- G06F16/68—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/683—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/60—Information retrieval; Database structures therefor; File system structures therefor of audio data
- G06F16/68—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/686—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using information manually generated, e.g. tags, keywords, comments, title or artist information, time, location or usage information, user ratings
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10H—ELECTROPHONIC MUSICAL INSTRUMENTS; INSTRUMENTS IN WHICH THE TONES ARE GENERATED BY ELECTROMECHANICAL MEANS OR ELECTRONIC GENERATORS, OR IN WHICH THE TONES ARE SYNTHESISED FROM A DATA STORE
- G10H2210/00—Aspects or methods of musical processing having intrinsic musical character, i.e. involving musical theory or musical parameters or relying on musical knowledge, as applied in electrophonic musical tools or instruments
- G10H2210/031—Musical analysis, i.e. isolation, extraction or identification of musical elements or musical parameters from a raw acoustic signal or from an encoded audio signal
- G10H2210/036—Musical analysis, i.e. isolation, extraction or identification of musical elements or musical parameters from a raw acoustic signal or from an encoded audio signal of musical genre, i.e. analysing the style of musical pieces, usually for selection, filtering or classification
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10H—ELECTROPHONIC MUSICAL INSTRUMENTS; INSTRUMENTS IN WHICH THE TONES ARE GENERATED BY ELECTROMECHANICAL MEANS OR ELECTRONIC GENERATORS, OR IN WHICH THE TONES ARE SYNTHESISED FROM A DATA STORE
- G10H2240/00—Data organisation or data communication aspects, specifically adapted for electrophonic musical tools or instruments
- G10H2240/095—Identification code, e.g. ISWC for musical works; Identification dataset
- G10H2240/101—User identification
- G10H2240/105—User profile, i.e. data about the user, e.g. for user settings or user preferences
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10H—ELECTROPHONIC MUSICAL INSTRUMENTS; INSTRUMENTS IN WHICH THE TONES ARE GENERATED BY ELECTROMECHANICAL MEANS OR ELECTRONIC GENERATORS, OR IN WHICH THE TONES ARE SYNTHESISED FROM A DATA STORE
- G10H2250/00—Aspects of algorithms or signal processing methods without intrinsic musical character, yet specifically adapted for or used in electrophonic musical processing
- G10H2250/131—Mathematical functions for musical analysis, processing, synthesis or composition
- G10H2250/215—Transforms, i.e. mathematical transforms into domains appropriate for musical signal processing, coding or compression
- G10H2250/235—Fourier transform; Discrete Fourier Transform [DFT]; Fast Fourier Transform [FFT]
Definitions
- the invention disclosed herein relates generally to methods and systems for analyzing and using digital music compositions, and more particularly to a method and system for determining the characteristics of a musical composition by analyzing its digital composition, and recommending particular musical compositions to users based upon the relative comparability of a user's desired musical characteristics and the musical characteristics of a collection of digital music.
- an artificial intelligence song/music recommendation system and method that allows music shoppers to discover new music.
- the system and method accomplish these tasks by analyzing a database of music in order to identify key similarities between different pieces of music, and then recommends pieces of music to a user depending upon their music preferences.
- a system and method is provided to compare specific new songs to a database of music to establish a score according to the new song's similarities with current ‘hit’ songs.
- a user can provide a digital music file of an original song that can be analyzed and compared to a ‘hit’ song database.
- the system provides a numerical score depending upon mathematical similarities to songs in the current ‘hit’ song database.
- a “Music Taste Test” function learns a user's music preferences via a series of binary choice questions, and delivers lists and/or personalized song recommendations to the user based on this information. Recommendations are prioritized and listed in order of closest song match on a theoretical multi-dimensional grid.
- a “More Like This” contextual matching function delivers song recommendations to the user based on a specific song as the defined starting point, and delivers songs that are most similar.
- a “My Personal Preferences” taste matching function allows the user to select and rate songs from the overall database of songs in the system, and personalized recommendations are received based on the personalized ratings.
- the system analyzes a music data file with extra parameters relating to commercial success of the music in the database. Using this extra dimension, new songs can be compared with the database to see how well such new song fits into the current market and to identify potential hits.
- FIG. 1 is a schematic view of a system architecture for the system of the instant invention.
- FIG. 2 is an exemplary single server configuration using the system of the instant invention.
- FIG. 3 is an exemplary multiple server configuration using the system of the instant invention.
- FIG. 4 is a view of a portion of the universe of analyzed songs illustrating some hit clusters.
- FIG. 5 is a view of a portion of the universe of analyzed songs shown in FIG. 4 , with a collection of new songs superimposed thereon.
- the method and system of the instant invention enable the analysis and processing of digital music in order to establish a description of a number of characteristics of the music, and likewise enable recommending a collection of music having particular characteristics to a user who has established a desired musical characteristic profile.
- the method and system of the instant invention enable identification of new songs that have similar characteristics with songs already established as commercial successes.
- the first step performed by the system is to analyze an existing digital music file in order to create a descriptive profile for the musical characteristics of the song.
- a database containing a library of digital music is provided.
- Such music database may be a part of the system employing the other functions described herein, or may be a remote resource that is maintained by a third party digital music company through remote, secure access. Access to such a third party digital music library may be provided, assigned, and protected under terms established by the particular third party database provider using known technology.
- the database preferably includes digital music files formatted as uncompressed linear PCM (“Pulse Code Modulation”) audio data. If stored in the database in a compressed format, the digital music files are decompressed using a standard tool for the compression scheme in use.
- the database may be modified, such as by adding additional digital music files to the database on a regular, periodic basis, such as weekly or monthly, to fit the particular business needs and objectives of the database provider.
- New analysis (as described herein) may also be done on a periodic basis similar to database updates, such that new music provided to the database may be included in the music recommendation functions described herein. Any newly generated analysis data from newly added digital music files is simply added at the end of the analysis data file, which is described in greater detail below.
- FIG. 1 depicts a schematic view of a system architecture for enabling the transfer and processing of digital music files to an automated digital music file analysis tool in order to generate an output file that, as described in greater detail below, serves as a descriptor of the musical characteristics of the particular musical composition that was analyzed.
- an “analyst wrapper” script receives digital music files from the music database or a song file archive, preferably copies the particular digital music file to a location on a computer network on which the analysis system is stored, decompresses the digital music file when necessary, and passes the linear PCM file to the analysis engine for processing.
- the analysis engine is labeled “HMI analyst” for Human Media Interface.
- the output of such analysis engine is directed to an output text file that, in turn, may be used by the recommendation utilities described below.
- all songs that are to be analyzed are processed in the same way, in series.
- the particular list of songs to be processed may vary depending upon the application. For instance, for a particular retailer, only their music catalogue is analyzed.
- all songs available in the digital music database are analyzed.
- the purpose of the initial analysis performed by the system is to analyze a variety of physical parameters of the music stored in the target digital music database. Such physical parameters describe quantifiable characteristics of music that may be mathematically modeled to create a descriptive, electronic “footprint” for each song.
- the analyzed parameters are based on human perception, and the system is referred to as a Human Media Interface (HMI) system.
- HMI Human Media Interface
- the system particularly analyzes one or more of the following characteristics for each musical composition: brightness, bandwidth, volume, tempo, rhythm, low frequency, noise, octave, and how these characteristics change over time, as well as length of the audio data. Not all of the characteristics necessarily provide distinctions in the music. Combinations of some or all of these characteristics may be employed without departing from the spirit and scope of the instant invention.
- the digital music file is divided into “chunks” that are separately processed in order to measure the characteristics for each such “chunk.” “Chunk” size is fixed and selected for optimizing performance over a test sample of songs so as to provide an appropriately representative sample for each parameter of interest.
- the value for each parameter in each chunk is measured. Parameters are measured over all “chunks” and averaged. The values that are derived through such measurements are, as described in greater detail below, used, in turn, to establish mean values and standard deviations among the parameter data so as to develop a profile for the entire song file.
- the data from the digital music file is decomposed using Fast Fourier Transform (FFT) techniques.
- FFT Fast Fourier Transform
- the Fourier transform in essence, decomposes or separates a waveform or function into FFT coefficients of different frequency that sum to the original waveform.
- the Fourier transform identifies or distinguishes the different frequency FFT coefficients and their respective amplitudes. Specific details concerning the application of FFT techniques can be found in P. Duhamel and M. Vetterli, Fast Fourier Transforms: A Tutorial Review And A State Of The Art, Chapter 19: Signal Processing, pp. 259-299 (Elsevier Science Publishers B. V. 1990), which is incorporated herein by reference thereto.
- each digital music file is decomposed into a defined set of FFT coefficients, which are grouped and separated in order to determine the different parameters to be calculated for each song.
- raw data in the form of the digital music file in PCM format, is directed to a script that decompresses the file (where necessary), divides the data into “chunks,” and applies FFT techniques to the data in each “chunk” to, in turn, establish the FFT coefficients for each “chunk.” Once those coefficients are established for each “chunk,” particular coefficients are chosen in order to calculate a numerical value for each of the parameters of interest in each “chunk.” More particularly, the specific parameters for each chunk are calculated, using the FFT coefficients, based on the following principles:
- parameter vector is formatted as a text file (with values separated by semicolons).
- An exemplary parameter vector may appear as follows:
- Negative and “too high” parameters correspond to out layers in the standardized distribution of frequencies of each parameter.
- the cut-offs are defined over a number of standard deviations calculated in a test sample of songs.
- the analysis tool analyzes files assuming that they are standard digital music files and assumes all data is valid (unless negative values are found). Such digital music files are typically identified with the filename extension .wav(.)
- the analysis tool reads audio waves of a song and performs algorithmic analyses of certain subjective criteria, as described above.
- the analysis tool divides the entire song into small sections (on the order of several seconds, depending upon the precision required), and calculates all of the parameter values for each of those sections using the same FFT analysis methods set forth above.
- the resulting parameter vectors thus represent the instantaneous value for all of the parameters at the particular song position specified by each section.
- each parameter vector for each section is compared to the parameter vector derived from the analysis of the entire song.
- the closest section's parameter vector is chosen as the most representative part of the song. “Closest” is defined in terms of unweighted quadratic differences between each section and the whole song. The parameters are already “equalized” by the standardization procedure.
- the system produces a song clip that includes that most representative part. In order to create a clip of a particular desired length (X), the system subtracts X/2 seconds from the position of the most representative section's position in order to obtain a clip of length X having the most representative part in the middle of the clip.
- the output from the analysis process described above for the complete song is stored in a plain text format, tab-delimited.
- the next step following analysis and profiling of the library of songs is to provide a recommendation engine to enable the system to recommend songs to particular users based upon the musical characteristics of the songs as depicted by their parameter vectors.
- recommendation functionality may be provided as an “in store” application running on the same server as the analysis component; on a server at the same location as the database of song parameter vectors, as illustrated in FIG. 2 ; on a remote server in communication with an application server at a retail location across a TCP/IP connection, as illustrated in FIG. 3 ; or the like.
- the recommendation engine loads the parameter vector file, and is configured to particularly deal with the text format of the parameter vector file.
- the text data of the parameter vector file may be loaded during the recommendation engine startup by specifying so on the command line, or a command may be sent after startup to locate and load the data.
- Data is normalized and centered by the system using Zero Mean Data and Standard Deviation Equal to Variance Equal to 1 for each parameter separately. More particularly, in applying Zero Mean Data, the mean value for each of the parameters is calculated across the entire song database, and the individual parameters of individual songs are then centered using those means. In other words, the parameter values for each song are changed by subtracting the calculated mean for each parameter by the original parameter value of the song. Thus, if the mean was recalculated with these new values, the mean value would be zero. Further, standard deviation and variance are correlated, as variance equals the square of the standard deviation. In applying Standard Deviation Equal to Variance Equal to 1, the data is normalized using standard deviation equal to 1.
- the system calculates the standard deviation and the variance for the set, and then, changing all of the values for the set (dividing the Zero Mean Data already calculated by the variance of the set) in order to get standard deviation equal to 1 if it was recalculated using these new values.
- Data correlations are determined using a standard correlation matrix.
- V ij is the value of the intersection of column i and row j
- M is the number of songs in the database
- Z i and Z j are parameter number i and j respectively for each song c
- the correlation matrix multiplies, for each pair of parameters, all values for all registers of data, adding them, and then dividing by the total number of registers.
- the correlation matrix is then used to calculate Eigenvectors and Eigenvalues.
- the correlation matrix is symmetric, standard Jacobi transformations (of symmetric matrices) are used to compute all Eigenvectors and Eigenvalues.
- Eigenvectors and Eigenvalues are standard statistical formulae used to determine and eliminate correlations between parameters, thereby reducing repeated information and increasing data utility.
- N is the number of parameters, and also, the number of eigenvectors
- E ij (where i and j go from 1 to N) are the values for parameter j of Eigenvector i
- Eigenvalues give the amount of uncorrelated information contained in each parameter. Thus, these values are used to weight the importance of parameters (or even discard parameters that contain no or almost no information).
- weighting and discarding rules are set for each customer. Weights for various parameters are session-dependent and user-dependent. For example, in a three parameters example, the parameters P 1 , P 2 and P 3 contain 75%, 23%, and 2% of the total amount of information, respectively. Depending on the specific customer needs (speed, accuracy, simplicity, etc.) the following weighting rules can be applied:
- the parameters are now uncorrelated, normalized, and orthogonal and weights have been set for each one, so the system can get maximum value from them. These new parameters replace, in the system, the old ones that came from the music analysis tool.
- Standard clustering algorithms are then employed by the system to locate and determine clusters within the entire database for further usage (specific to music recommendation functions “More Like This” and “My Personal Preferences,” described in greater detail below).
- the system may provide a variety of music recommendation functions through interaction with a user.
- One method is to capture a user's own personal taste profile by enabling them to take a “music taste test”. In this process, the user is presented with a number of binary choices between two short audio clips and chooses the clip they prefer. After a series of comparisons, it is possible to generate a profile for that user; the profile is analogous to a song's own profile, as measured in the analysis phase. In this way, songs from the database that share commonalities to the user profile can be identified and presented to the user to preview. The second way simply links a song or a group of songs to a selection of music that has a similar profile. This technique takes the individual profile of the song or songs and matches it to the whole catalogue of music in the database.
- a user interacts with a computer, website, interactive kiosk, in-store listening station (like those commercially available from ACD, Fullplay, Black Box AV, First Universal, and RedDotNet), or other computer-like device in which the recommendation functions are loaded.
- a computer website, interactive kiosk, in-store listening station (like those commercially available from ACD, Fullplay, Black Box AV, First Universal, and RedDotNet), or other computer-like device in which the recommendation functions are loaded.
- FIGS. 2 and 3 user interaction takes place at one of the client devices, as shown.
- the recommendation software can be loaded directly to the computer/device/server, as shown in the single server configuration of FIG. 2 .
- the recommendation software can be served remotely through a LAN, WAN, or via an ASP environment, as shown in the multiple server configuration of FIG. 3 .
- the music recommendation functions are preferably executed as a series of requests and responses handled by a daemon (or service) on a server.
- the daemon listens for requests to its network address on a specific port, and communicates with client devices using its own protocol, referred to herein as Simple HMI-Music Protocol (SHMP).
- SHMP is a protocol designed to satisfy the functionality given by the system over any TCP/IP network, and is used to send and receive a variety of requests, including, but not exclusive to, “Music Taste Test Request” and “More Like This Song,” as described in greater detail below.
- a “Music Taste Test” (“MTT”) is used to determine a user's music preferences, and thereby make personalized song recommendations.
- the MTT function is implemented through a pattern-recognizing artificial intelligence system using an MTT daemon (labeled ‘mttd’ in FIGS. 2 and 3 ).
- the MTT function creates simple questions and instructions that prompt a user to listen to a series of two short music clips and make a choice on which music clips he/she prefers between the two choices.
- the first two clips are initially selected at random to maximize the dissimilarity in the parameters corresponding to each song.
- Subsequent selections are performed by using the Eigenvector algorithm, maximizing the information given by the selection.
- the Eigenvector algorithm eliminates correlated data that cannot be used to distinguish between the two songs.
- the system recognizes which parameter(s) provide distinguishing characteristics and uses the user input to select another two choices in order to capture the taste of the user regarding all of the variable parameters.
- the sequence of binary choices is repeated until the MTT successfully defines preferred values for each parameter or if the system is prompted to give recommendations. Since the system learns from every user response, it can recommend appropriate songs at any point in the test with the information it has learned. In order to do this, the MTT function asks questions based upon the different parameters and decreases the size of the preferred subset according to the songs chosen by the user. Essentially, the system provides the representative clip from two songs previously analyzed and asks the user to select which song the user prefers. By asking binary questions, the system finds the preferred values for each one of the song parameters analyzed.
- the preferred range for each parameter gets smaller too and, at some point, the user will not be able to distinguish between the two extremes of the parameter for the range found in the subset. In other words, the user will equally like songs from the whole range.
- the system is able to tell the user's perception level for each parameter and determine that the system cannot learn any more from the user.
- the MTT function identifies that it has reached this point for a parameter when the user chooses songs indistinctly from both extremes of the subset's range, and so there is no determined consistency in the answers.
- the MTT function will never have a 100% confidence level for a user (as long as there remain songs in the database); thus, in reality, a new pair of songs can always be requested.
- the MTT function's learning process seeks to achieve the highest possible confidence level for all parameters, in essence achieving a high confidence level for the user's preferences.
- the system seeks to reach a high confidence level for each user and develops a precise taste profile for each user (based on the parameters analyzed).
- this process can continue endlessly, as the MTT function seeks to shorten preferred ranges and modify preferred values.
- it is necessary to set a limit on the MTT process. This limit can be set either by setting a limited number of sound clip comparisons, target confidence level, or both. This is a matter of simple programming on the front-end application on the customer's side.
- a user's profile is stored in the system while the user is still connected to the system. When the session is terminated, the user profile is eliminated. However, this profile can also be saved within the system or external database and accessed at a later date, depending upon the specific retail customer and implementation environment. Thus, a user can go back, continue the MTT process, and obtain refined recommendations. This functionality is achieved using an identification technology such as loyalty cards or cookies, as described in greater detail below.
- the MTT process in a retail situation is adapted to require approximately 8-15 steps to complete the taste test determination process.
- a progress meter is optionally displayed on the user's screen reflecting the confidence level of the system, thereby allowing users to understand where they are in the test process.
- the typical user interface includes two basic “play buttons” that allow the user to first preview both music clips (multiple times, if they desire), and two basic “select buttons” that then allow the user to select which of the two music clips they prefer. There is no qualitative aspect to these choices; rather, the user simply decides which song is preferred between the two music clips.
- the music clips used in the MTT process are unique audio files, derived from the clip extraction phase described in detail above.
- the MTT user interface may vary based on implementation environments and operating systems.
- the MTT functionality is independent of the user interface as long as the user is presented with the two songs and is able to choose the one that he/she prefers.
- a user interface could consist of two songs displayed as text and then the user is prompted to type the name of his preferred song, the MTT function will continue to function the same.
- the interface can be customized to meet the specific needs/requirements of different customers. This includes adding a customer's logo, color scheme, wording/tonality, and audio cues.
- the user's particular musical tastes are established in the user's personal ‘taste’ vector.
- the system uses the ‘taste’ vector to determine and display a list of song matches to the user's determined musical taste. Alternately, the list of song matches may be displayed to the user at his or her request before completion of the analysis, in which case the recommendations will simply reflect the MTT function's current confidence level.
- the MTT function searches the music database in order to find songs that most closely match that user's ‘taste’ profile. Closeness of a song to the ‘taste’ vector is determined by the mathematical distance between two songs.
- the theory of the recommendation engine is based upon the relative correlation between a user's preferred values and each song (where correlation is defined as the relative distance measured as the sum of the squared difference between each parameter).
- a song is always considered as a whole and, thus, individual parameters are not considered to be matches by themselves.
- the “More Like This” (“MLT”) function is implemented through a pattern-recognizing artificial intelligence system using an MLT daemon (labeled ‘mltd’ in FIGS. 2 and 3 ).
- MLT MLT daemon
- the MLT function allows a user to receive music recommendations by selecting a song and requesting songs that are mathematically similar to that song.
- the user needs to enter the name of a song into the device or simply click on the “More Like This” button to receive recommendations.
- the system When the MLT function is operable, the system performs real-time MLT recommendations by looking for song files that have the most mathematically similar analysis data to the parent song.
- the MLT lookup process consists of searching for similar songs by checking the relative distances between all parameters for each song. It is the difference between parameters that determine if the songs are similar or not for the user. Therefore, given a list of songs, each song can have a “More Like This” link to similar music.
- a “My Personal Preferences” (“MPP”) function enables a user to establish a personal preference file that may in turn be used by the system to provide the user with music selections that match that user's personal profile.
- MPP My Personal Preferences
- a user While interacting with the device or website, a user ranks songs against preferably a simple five level scale (e.g., “I love it”, “I like it”, “I neither like nor dislike it”, “I don't like it”, and “I strongly dislike it”) by choosing the appropriate option on a user interface screen. This process can be done individually for each song listened to, or by entering song names and their rankings into the device. The user can add, erase, or edit songs and rankings from their list at any time. When requested, recommendations can be made using the current songs and ratings on the list.
- the parameter vectors for the MPP selected songs are weighted, based on a user's rankings, and are input to the system.
- This process is similar to the MTT learning process except, instead of the system deciding what it wants to learn and asking the appropriate questions, the user chooses the songs (thereby choosing the parameter vectors) and forces the system to learn from the user inputs. This is referred to as forced learning.
- Selected songs are grouped based on the given rank in order to find preference groups.
- the selected songs fall into two types, liked songs and disliked songs. Songs ranked as “I love it”, “I like it”, or “I neither like nor dislike it” are defined as liked songs, and songs ranked as “I don't like it” and “I strongly dislike it” are defined as disliked songs. Similar groupings of songs are important to a successful MPP process.
- Grouping is accomplished by searching through the ranked songs to find commonalities within each type (liked and disliked songs), and then grouping the songs that are most similar (i.e. close in proximity to each other). Note that a person can have more than one preference grouping within each preference group type (e.g. someone can both like slow music and hard rap music). Thus, preference groups are also referred to as musical tastes (and are also either liked or disliked).
- Grouping in this way is also done to eliminate songs that are ranked by the user, but dissimilar to other songs of the same type. Songs are eliminated in this way because they are not valuable to the MPP recommendation process (i.e., the system cannot determine what it is that the user likes from that song).
- Mean values are calculated for each musical taste to determine the corresponding parameter vector (referred to as a musical taste vector).
- songs within each musical taste are weighted according to the user's ranking for those songs, thus giving less importance to the “I like it” songs than to the “I love it” songs, but more than to the “I neither like nor dislike it” songs, for the liked musical tastes; and giving more importance to the “I strongly dislike it” than to the “I don't like it”, for the disliked musical tastes.
- the system biases the mean towards the more liked songs and the more disliked songs.
- the system is ready to recommend songs to the user.
- the system needs to create a list of possible songs to be recommended.
- the initial list contains all songs in the database.
- songs that are known to be disliked by the user are erased by analyzing in which relative clusters (within the whole database of songs) the disliked musical taste vectors fall (see discussion above concerning elimination of analysis data correlations). All songs from the clusters containing disliked musical taste vectors are erased from the list of possible songs to be recommended.
- recommendations are generated from the liked musical taste vectors using only the list of possible songs to be recommended.
- the system can either present the musical tastes found to the user (from which to select their recommendations) or simply deliver a combined recommendation list including songs from all different musical tastes.
- both the MTT and MLT functions also create user tracking statistics in the form of log files.
- log file information is of a similar format to most daemon software and includes information as discrete calls to the recommendation engine including session number, the device requesting the action, date and time, and the outcome of the action. Every call creates a log entry and thus, unique user sessions within the Human Media Interface system can be tracked and monitored.
- An example log excerpt is as follows:
- HELO 8 23 gives both the third party's id number: 8 (e.g. the retailer) and the terminal id (the actual in-store device that is requesting this transaction: 23).
- Anonymous user session data (not specific to an individual) can be used to determine general trends within the system and therefore is valuable from a business/marketing perspective.
- Examples of anonymous user session data include number of users per time period, user paths through the system, most/least recommended music, error messages, and average length of time per session.
- User session information can also be attributed to an individual (e.g. John Smith logged into music kiosk #2 at 4:30 PM on Aug. 23, 2002). This can be done via the use of scanable loyalty cards or cookies in a web environment.
- a loyalty card or customer card in this case is a card-like physical identifier, such as those currently used by several retail chains and supermarkets.
- the user On arrival to a kiosk or in-store music device, the user is prompted to scan their card, thereby enabling the system to associate a specific user session to a person. Online this is done via “cookies”, as is commonly used on the Internet to identify users.
- Associated user session data is more robust, as various sessions from the same user can be linked together and personal information garnered.
- the core of this personal information includes but is not limited to personal music preferences (as deemed by the MTT and MLT functions). This is extremely useful business information for personalized sales and marketing efforts. An example of this includes presenting an individual a personalized music coupon at checkout, offering a CD from which they have already previewed in the system.
- personalized user session data information can be combined with other marketing databases and exported to various CRM systems, thereby enabling an even higher level of business intelligence.
- a retailer can use their unified CRM system to combine an individual's system-derived music preferences, select a CD to be recommended to the individual, check current inventory, determine dynamic pricing for the CD, and deliver a personalized offer via email.
- the technology of the present invention can be used to predict the potential commercial success of a new song.
- a large database is analyzed to determine a song vector for each song, as described above.
- approximately 3.5 million songs including almost everything that has been released by the music labels since the 1950's until the present time has been analyzed.
- the database is updated weekly with new releases.
- Each song is then mapped onto a grid called the music universe and is positioned according to its mathematical characteristics.
- Each song is represented by a dot on the universe and the songs on one end of the universe are vastly different from songs on the other end of the universe.
- Songs with mathematical similarities are positioned very close to one another.
- all of the songs from the music universe that had not been hits in the past five years were removed. The songs remaining were grouped into a limited number of small clusters all over the universe but with vast spaces between them, as illustrated in FIG. 4 .
- “Hit” songs seem to have common characteristics based on the mathematical analysis. While other factors, such as lyrics, theme and artist, influence the success of a song, it is rare for a song that falls outside of the clusters to become a hit.
- the present invention takes the analyzed song data for the entire database and overlays extra parameters relating to the commercial success of the music in the database. These parameters may include additional data such as total sales, highest chart position, date of release, and other common indicators of commercial success.
- the specific weight given to such new parameters can be varied to meet an end users specific needs.
- new releases, potential releases, and even unsigned acts can be compared with the database to see how well a given song fits into the current market and to identify potential hits. As the market changes, the system reflects such changes by finding new patterns in the hit clusters and applying these to the process.
- the system allows for trends to be identified as they develop over time, meaning that a song that contains strong characteristics that are becoming more prevalent in new music and less of the characteristics that are diminishing can be identified as having high potential. Simply put this means that a song that sounds uncommercial to a human listening to it now may just be ahead of its time and in fact contains the right ingredients to appeal to the public sometime in the future.
- a database of past chart hits from either the US or the UK is used, as well as a database of “classic hits” from 1956 to the present time as the universe of hits.
- the “hit” database in an archive of every song that has been in the US Weekly Top 30 or the UK Official Charts for the past 5 years. Songs are weighted based upon their ranking within the Top 30. As new songs are added, old songs are deleted. This is done to keep the database as fresh as possible, effectively following the trends and styles as music changes over time.
- a user can submit one, or more, digital music files for analysis.
- the system scores a new song according to its similarities with current hit songs.
- a parameter vector is determined for the new song.
- the patterns in such new song are compared to patterns in recent chart “hits” and to patterns in classic hit songs going back to 1956.
- the new song is compared to the “hits,” by comparing the new song parameter vector to the “hit” parameter vector, as described above, to obtain an affinity value for each song in the “hit” database.
- the affinity value is a rating that shows how closely related the mathematical patterns in one song are to another. The lower the affinity value between two songs the more closely related they are.
- An affinity value under 1.00 denotes close similarity while an affinity value of over 1.00 denotes similarity but less obvious and more distant similarity.
- FIG. 5 is an illustration of 12 new songs as analyzed and graphed on the universe of “hits”. In this illustration, there are two songs that fall squarely within hit clusters.
- the system gives the new song a score.
- an overall score is assigned to the new song.
- a rating of 7.00 or greater should be considered to be a very strong song and very closely related to what is currently popular in the market.
- the system can give the new song a recentness score and a sales score.
- a report is provided to the user who submitted the new song indicating the overall score, the proximity to “hit” songs, and the proximity to current “hit” songs. Using such information, the user can determine an appropriate course of action for commercializing the analyzed new song.
Abstract
An artificial intelligence song/music recommendation system and method is provided that allows music shoppers to discover new music. The system and method accomplish these tasks by analyzing a database of music in order to identify key similarities between different pieces of music, and then recommends pieces of music to a user depending upon their music preferences. An embodiment enables a user to evaluate a new song's similarity to songs already established as commercially valuable.
Description
- This application is a continuation of co-pending and co-owned U.S. patent application Ser. No. 10/678,505, filed with the U.S. Patent and Trademark Office on Oct. 3, 2003 entitled “Method and System for Music Recommendation”, now U.S. Pat. No. 7,081,579, which is based upon and claims benefit of copending and co-owned U.S. Provisional Patent Application Ser. No. 60/415,868 entitled “Method and System for Music Recommendation”, filed with the U.S. Patent and Trademark Office on Oct. 3, 2002 by the inventors herein, the specifications of which are incorporated herein by reference.
- 1. Field of the Invention
- The invention disclosed herein relates generally to methods and systems for analyzing and using digital music compositions, and more particularly to a method and system for determining the characteristics of a musical composition by analyzing its digital composition, and recommending particular musical compositions to users based upon the relative comparability of a user's desired musical characteristics and the musical characteristics of a collection of digital music.
- 2. Background of the Invention
- Historically, what is pleasing to the human ear has not changed since man began making sounds. Patterns in music that are pleasing to the human ear have not changed much, if at all, since the times of the classical composers. What has changed are styles, performances, the instruments used, and the way music is produced and recorded, but a compelling melody is still compelling and a series of random notes still sounds random. For example, the dictionary describes melody as a series of notes strung together in a meaningful sequence. Unfortunately, some sequences sound meaningful and make up a beautiful song and other sequences just sound like noise.
- While the number of possible melody patterns combined with all of the other variables in recorded music allow for a seemingly infinite number of combinations, the patterns that we find pleasing have not changed. That is not to say everything has been invented, however. So far, every new style of music that has come into being: country, rock, punk, grunge etc. have all had similar mathematical patterns. The hits in those genres have all come from the same ‘hit’ clusters that exist today and anything that has fallen outside of such ‘hit’ clusters has rarely been successfully on the charts for its musical qualities.
- It is an object of the present invention to provide a method and system for measuring the characteristics of a musical composition, and establishing a collection of digital musical compositions that may be sorted based upon such characteristics.
- It is another object of the present invention to provide a method and system for determining a preferred musical characteristic profile for a music listener.
- It is another object of the present invention to enable a method and system to compare new digital music files with historical commercially successful songs.
- In accordance with the above objects, an artificial intelligence song/music recommendation system and method is provided that allows music shoppers to discover new music. The system and method accomplish these tasks by analyzing a database of music in order to identify key similarities between different pieces of music, and then recommends pieces of music to a user depending upon their music preferences.
- In an alternate embodiment, a system and method is provided to compare specific new songs to a database of music to establish a score according to the new song's similarities with current ‘hit’ songs. A user can provide a digital music file of an original song that can be analyzed and compared to a ‘hit’ song database. The system provides a numerical score depending upon mathematical similarities to songs in the current ‘hit’ song database.
- To provide users with music recommendations, the system employs a number of analysis functions. First, a “Music Taste Test” function learns a user's music preferences via a series of binary choice questions, and delivers lists and/or personalized song recommendations to the user based on this information. Recommendations are prioritized and listed in order of closest song match on a theoretical multi-dimensional grid. Next, a “More Like This” contextual matching function delivers song recommendations to the user based on a specific song as the defined starting point, and delivers songs that are most similar. Lastly, a “My Personal Preferences” taste matching function allows the user to select and rate songs from the overall database of songs in the system, and personalized recommendations are received based on the personalized ratings.
- To provide users with an analysis of an original song, the system analyzes a music data file with extra parameters relating to commercial success of the music in the database. Using this extra dimension, new songs can be compared with the database to see how well such new song fits into the current market and to identify potential hits.
- Other objects, features, and advantages of the present invention will become more apparent from the following detailed description of the preferred embodiments and certain modifications thereof when taken together with the accompanying drawings, in which:
-
FIG. 1 is a schematic view of a system architecture for the system of the instant invention. -
FIG. 2 is an exemplary single server configuration using the system of the instant invention. -
FIG. 3 is an exemplary multiple server configuration using the system of the instant invention. -
FIG. 4 is a view of a portion of the universe of analyzed songs illustrating some hit clusters. -
FIG. 5 is a view of a portion of the universe of analyzed songs shown inFIG. 4 , with a collection of new songs superimposed thereon. - The invention summarized above and defined by the enumerated claims may be better understood by referring to the following description, which should be read in conjunction with the accompanying drawings. This description of an embodiment, set out below to enable one to build and use an implementation of the invention, is not intended to limit the enumerated claims, but to serve as particular examples thereof. Those skilled in the art should appreciate that they may readily use the conception and specific embodiments disclosed as a basis for modifying or designing other methods and systems for carrying out the same purposes of the present invention. Those skilled in the art should also realize that such equivalent assemblies do not depart from the spirit and scope of the invention in its broadest form.
- The method and system of the instant invention enable the analysis and processing of digital music in order to establish a description of a number of characteristics of the music, and likewise enable recommending a collection of music having particular characteristics to a user who has established a desired musical characteristic profile. In an alternate embodiment, the method and system of the instant invention enable identification of new songs that have similar characteristics with songs already established as commercial successes.
- The first step performed by the system is to analyze an existing digital music file in order to create a descriptive profile for the musical characteristics of the song. Preferably, a database containing a library of digital music is provided. Such music database may be a part of the system employing the other functions described herein, or may be a remote resource that is maintained by a third party digital music company through remote, secure access. Access to such a third party digital music library may be provided, assigned, and protected under terms established by the particular third party database provider using known technology.
- Whether provided as an element of the analysis and recommendation system or as a third party remote resource, the database preferably includes digital music files formatted as uncompressed linear PCM (“Pulse Code Modulation”) audio data. If stored in the database in a compressed format, the digital music files are decompressed using a standard tool for the compression scheme in use. The database may be modified, such as by adding additional digital music files to the database on a regular, periodic basis, such as weekly or monthly, to fit the particular business needs and objectives of the database provider. New analysis (as described herein) may also be done on a periodic basis similar to database updates, such that new music provided to the database may be included in the music recommendation functions described herein. Any newly generated analysis data from newly added digital music files is simply added at the end of the analysis data file, which is described in greater detail below.
-
FIG. 1 depicts a schematic view of a system architecture for enabling the transfer and processing of digital music files to an automated digital music file analysis tool in order to generate an output file that, as described in greater detail below, serves as a descriptor of the musical characteristics of the particular musical composition that was analyzed. As shown inFIG. 1 , an “analyst wrapper” script receives digital music files from the music database or a song file archive, preferably copies the particular digital music file to a location on a computer network on which the analysis system is stored, decompresses the digital music file when necessary, and passes the linear PCM file to the analysis engine for processing. (InFIG. 1 , the analysis engine is labeled “HMI analyst” for Human Media Interface.) The output of such analysis engine is directed to an output text file that, in turn, may be used by the recommendation utilities described below. - During the initial analysis performed by the system, all songs that are to be analyzed are processed in the same way, in series. The particular list of songs to be processed may vary depending upon the application. For instance, for a particular retailer, only their music catalogue is analyzed. For a generalized music recommendation system, all songs available in the digital music database are analyzed. The purpose of the initial analysis performed by the system is to analyze a variety of physical parameters of the music stored in the target digital music database. Such physical parameters describe quantifiable characteristics of music that may be mathematically modeled to create a descriptive, electronic “footprint” for each song. Moreover, the analyzed parameters are based on human perception, and the system is referred to as a Human Media Interface (HMI) system. The characteristics have been identified to be the ones that produced the strongest reaction in testers. Often the listener detects the characteristics unconsciously. In general, the mix of parameters is more important than any individual parameter. To implement the methods described herein, the system particularly analyzes one or more of the following characteristics for each musical composition: brightness, bandwidth, volume, tempo, rhythm, low frequency, noise, octave, and how these characteristics change over time, as well as length of the audio data. Not all of the characteristics necessarily provide distinctions in the music. Combinations of some or all of these characteristics may be employed without departing from the spirit and scope of the instant invention.
- In order to measure each of these characteristics, the digital music file is divided into “chunks” that are separately processed in order to measure the characteristics for each such “chunk.” “Chunk” size is fixed and selected for optimizing performance over a test sample of songs so as to provide an appropriately representative sample for each parameter of interest. Once the data from the digital music file is divided into such “chunks,” the value for each parameter in each chunk is measured. Parameters are measured over all “chunks” and averaged. The values that are derived through such measurements are, as described in greater detail below, used, in turn, to establish mean values and standard deviations among the parameter data so as to develop a profile for the entire song file.
- In order to measure the parameters noted above, the data from the digital music file is decomposed using Fast Fourier Transform (FFT) techniques. The Fourier transform, in essence, decomposes or separates a waveform or function into FFT coefficients of different frequency that sum to the original waveform. The Fourier transform identifies or distinguishes the different frequency FFT coefficients and their respective amplitudes. Specific details concerning the application of FFT techniques can be found in P. Duhamel and M. Vetterli, Fast Fourier Transforms: A Tutorial Review And A State Of The Art, Chapter 19: Signal Processing, pp. 259-299 (Elsevier Science Publishers B. V. 1990), which is incorporated herein by reference thereto.
- Using such FFT techniques, each digital music file is decomposed into a defined set of FFT coefficients, which are grouped and separated in order to determine the different parameters to be calculated for each song. To accomplish this task, raw data, in the form of the digital music file in PCM format, is directed to a script that decompresses the file (where necessary), divides the data into “chunks,” and applies FFT techniques to the data in each “chunk” to, in turn, establish the FFT coefficients for each “chunk.” Once those coefficients are established for each “chunk,” particular coefficients are chosen in order to calculate a numerical value for each of the parameters of interest in each “chunk.” More particularly, the specific parameters for each chunk are calculated, using the FFT coefficients, based on the following principles:
-
- Brightness is quantified as a spectral histogram built from the FFT.
- Bandwidth is quantified as the variance of the spectral histogram built from the FFT.
- Tempo is quantified as a measure obtained using a “Beat Tracker” algorithm (based upon autocorrelation values). The Beat tracker algorithm calculates how many beats per second are in a “chunk.”
- Volume is quantified as the average change in the bit sample amplitudes at
lag 1.Lag 1 refers to consecutive beats. - Low Frequency is quantified as spectral power weighted with ten inverse frequency.
- Noise is quantified as the sum of all the FFT coefficients normalized to the maximum FFT coefficient.
- Octave is quantified as the temporal mean of Cepstrum (inverse of the logarithmic FFT). The Octave is calculated for each chunk and averaged.
- File Size is quantified as the number of bytes of the file.
In other words, FFT coefficients provide data needed to calculate all the parameters listed, except for file size. Thus, the result of the analysis of the data from the digital music file is several parameter values based on the number of particular “chunks.” In order to derive a final parameter value for the song as a whole, the mean is taken from all “chunks” within the same parameter.
- Compiling each value for all the parameters measured results in an output that is referred to herein as a “parameter vector,” which parameter vector is formatted as a text file (with values separated by semicolons). An exemplary parameter vector may appear as follows:
-
- 78221904926;1;01;41.018945;26.088065;38.450639;23.616234;5337.3284;266.8 48675;118.7692;76.484095;116.6154;73.836892;0.00007;0.00020;3;229.9736;31 0.573784;27.931500
- Note: The actual ordering of values in the set is immaterial so long as it is known and consistent.
The bold section of the sample parameter vector noted above is an example of song identification information, including the UPC, disk number, and track number. The other data is sample parameter value information (e.g., tempo, volume, etc.). Based upon the particular database, other generally available information may optionally be provided and may be useful in the music recommendation functions described in greater detail below. Such other information may include, by way of example: (i) retailer/store-specific media identification; (ii) set count (number of disks in set); (iii) album title; (iv) artist name; (v) music genre; (vi) record label; (vii) track title; (viii) track artist; (ix) track duration; and (x) copyright date.
- The system will recognize almost all values for any parameter as valid, except negative values. If negative values are found, it means that the digital music file is invalid, and the data needs to be manually discarded. Further, if values are consistently too high, there is the possibility of damaged data or data stemming from an invalid file. Negative and “too high” parameters correspond to out layers in the standardized distribution of frequencies of each parameter. The cut-offs are defined over a number of standard deviations calculated in a test sample of songs. The analysis tool analyzes files assuming that they are standard digital music files and assumes all data is valid (unless negative values are found). Such digital music files are typically identified with the filename extension .wav(.) The analysis tool reads audio waves of a song and performs algorithmic analyses of certain subjective criteria, as described above.
- Once the parameter vector for the entire song has been established as set forth above, standard mathematical analysis is used to extract samples of the most representative passages of the music. More particularly, the analysis tool divides the entire song into small sections (on the order of several seconds, depending upon the precision required), and calculates all of the parameter values for each of those sections using the same FFT analysis methods set forth above. The resulting parameter vectors thus represent the instantaneous value for all of the parameters at the particular song position specified by each section. When such section processing is completed, a parameter vector exists for each section.
- Following section processing, the system then proceeds to find the most representative passage (whose length varies depending upon the application) of the song. First, each parameter vector for each section is compared to the parameter vector derived from the analysis of the entire song. The closest section's parameter vector is chosen as the most representative part of the song. “Closest” is defined in terms of unweighted quadratic differences between each section and the whole song. The parameters are already “equalized” by the standardization procedure. Once the most representative part of the song is determined, the system produces a song clip that includes that most representative part. In order to create a clip of a particular desired length (X), the system subtracts X/2 seconds from the position of the most representative section's position in order to obtain a clip of length X having the most representative part in the middle of the clip.
- The output from the analysis process described above for the complete song is stored in a plain text format, tab-delimited. The next step following analysis and profiling of the library of songs is to provide a recommendation engine to enable the system to recommend songs to particular users based upon the musical characteristics of the songs as depicted by their parameter vectors. It should be noted that such recommendation functionality may be provided as an “in store” application running on the same server as the analysis component; on a server at the same location as the database of song parameter vectors, as illustrated in
FIG. 2 ; on a remote server in communication with an application server at a retail location across a TCP/IP connection, as illustrated inFIG. 3 ; or the like. In each instance, the recommendation engine loads the parameter vector file, and is configured to particularly deal with the text format of the parameter vector file. The text data of the parameter vector file may be loaded during the recommendation engine startup by specifying so on the command line, or a command may be sent after startup to locate and load the data. - Following the loading of the text data of the parameter vector to the recommendation engine, various mathematical and statistical procedures (discussed in detail below) are run on the loaded data to ensure that all of the data is meaningful, and to extract the essential characteristics from each song and its preferences profile. Such preferences profile is the parameter vector described above, after normalization has been performed based on the whole song analysis database (and is generated and formatted in the same manner).
- Some statistical methods are used to improve the quality of the analysis information and subsequent Artificial Intelligence techniques. Data is normalized and centered by the system using Zero Mean Data and Standard Deviation Equal to Variance Equal to 1 for each parameter separately. More particularly, in applying Zero Mean Data, the mean value for each of the parameters is calculated across the entire song database, and the individual parameters of individual songs are then centered using those means. In other words, the parameter values for each song are changed by subtracting the calculated mean for each parameter by the original parameter value of the song. Thus, if the mean was recalculated with these new values, the mean value would be zero. Further, standard deviation and variance are correlated, as variance equals the square of the standard deviation. In applying Standard Deviation Equal to Variance Equal to 1, the data is normalized using standard deviation equal to 1. To do this, the system calculates the standard deviation and the variance for the set, and then, changing all of the values for the set (dividing the Zero Mean Data already calculated by the variance of the set) in order to get standard deviation equal to 1 if it was recalculated using these new values.
- With centered and normalized data, correlations between all analysis parameters are identified and eliminated to reduce redundant information given by different parameters and getting the most unique information from each parameter. To do this, the system creates a new set of parameters with no correlations, but with all the information included in the parameters analyzed by the analysis tool. These new parameters have no conceptual meaning (i.e., they are referred to as parameter1, parameter2, etc., instead of “volume,” “tempo,” etc.).
- Data correlations are determined using a standard correlation matrix. The standard correlation matrix formula is as follows:
- Vij is the value of the intersection of column i and row j
- M is the number of songs in the database
- c is each particular song
- Zi and Zj are parameter number i and j respectively for each song c
- In essence, the correlation matrix multiplies, for each pair of parameters, all values for all registers of data, adding them, and then dividing by the total number of registers.
- The correlation matrix is then used to calculate Eigenvectors and Eigenvalues. As the correlation matrix is symmetric, standard Jacobi transformations (of symmetric matrices) are used to compute all Eigenvectors and Eigenvalues. Eigenvectors and Eigenvalues are standard statistical formulae used to determine and eliminate correlations between parameters, thereby reducing repeated information and increasing data utility.
- Eigenvectors are important in determining the new uncorrelated parameters from the original parameters, as illustrated in the following formula:
- Pi (where i goes from 1 to N) are the new parameter values for each song
- N is the number of parameters, and also, the number of eigenvectors
- Oi (where i goes from 1 to N) are the original parameter values for each song
- Eij (where i and j go from 1 to N) are the values for parameter j of Eigenvector i
- Eigenvalues give the amount of uncorrelated information contained in each parameter. Thus, these values are used to weight the importance of parameters (or even discard parameters that contain no or almost no information).
- Depending on precision and speed desired, weighting and discarding rules are set for each customer. Weights for various parameters are session-dependent and user-dependent. For example, in a three parameters example, the parameters P1, P2 and P3 contain 75%, 23%, and 2% of the total amount of information, respectively. Depending on the specific customer needs (speed, accuracy, simplicity, etc.) the following weighting rules can be applied:
- Keep all parameters equally weighted (33.33%, 33.33%, and 33.33%)
- Weight all parameters based on the percentages above (75%, 23%, and 2%)
- Discard the last parameter and weight the remaining two as P1 77% and P2 23%
- Discard the last parameter and equally weight the remaining two (50%, 50%)
- Etc.
- The parameters are now uncorrelated, normalized, and orthogonal and weights have been set for each one, so the system can get maximum value from them. These new parameters replace, in the system, the old ones that came from the music analysis tool.
- Standard clustering algorithms are then employed by the system to locate and determine clusters within the entire database for further usage (specific to music recommendation functions “More Like This” and “My Personal Preferences,” described in greater detail below).
- After a large selection of music has been analyzed and the data has been uncorrelated, normalized, orthogonal, and weights are set, the system may provide a variety of music recommendation functions through interaction with a user. There are two keys ways to recommend music to an individual user. One method is to capture a user's own personal taste profile by enabling them to take a “music taste test”. In this process, the user is presented with a number of binary choices between two short audio clips and chooses the clip they prefer. After a series of comparisons, it is possible to generate a profile for that user; the profile is analogous to a song's own profile, as measured in the analysis phase. In this way, songs from the database that share commonalities to the user profile can be identified and presented to the user to preview. The second way simply links a song or a group of songs to a selection of music that has a similar profile. This technique takes the individual profile of the song or songs and matches it to the whole catalogue of music in the database.
- In order to initiate the music recommendation features, a user interacts with a computer, website, interactive kiosk, in-store listening station (like those commercially available from ACD, Fullplay, Black Box AV, First Universal, and RedDotNet), or other computer-like device in which the recommendation functions are loaded. In
FIGS. 2 and 3 , user interaction takes place at one of the client devices, as shown. The recommendation software can be loaded directly to the computer/device/server, as shown in the single server configuration ofFIG. 2 . In other embodiments, the recommendation software can be served remotely through a LAN, WAN, or via an ASP environment, as shown in the multiple server configuration ofFIG. 3 . - Irrespective of the locality of the software, the music recommendation functions are preferably executed as a series of requests and responses handled by a daemon (or service) on a server. The daemon listens for requests to its network address on a specific port, and communicates with client devices using its own protocol, referred to herein as Simple HMI-Music Protocol (SHMP). SHMP is a protocol designed to satisfy the functionality given by the system over any TCP/IP network, and is used to send and receive a variety of requests, including, but not exclusive to, “Music Taste Test Request” and “More Like This Song,” as described in greater detail below.
- A “Music Taste Test” (“MTT”) is used to determine a user's music preferences, and thereby make personalized song recommendations. The MTT function is implemented through a pattern-recognizing artificial intelligence system using an MTT daemon (labeled ‘mttd’ in
FIGS. 2 and 3 ). The MTT function creates simple questions and instructions that prompt a user to listen to a series of two short music clips and make a choice on which music clips he/she prefers between the two choices. The first two clips are initially selected at random to maximize the dissimilarity in the parameters corresponding to each song. Subsequent selections are performed by using the Eigenvector algorithm, maximizing the information given by the selection. The Eigenvector algorithm eliminates correlated data that cannot be used to distinguish between the two songs. For example, if the numerical value for a selected parameter is substantially the same in the two songs, selection of either song by the user as being preferred is likely not due to the selected parameter. The system recognizes which parameter(s) provide distinguishing characteristics and uses the user input to select another two choices in order to capture the taste of the user regarding all of the variable parameters. - The sequence of binary choices is repeated until the MTT successfully defines preferred values for each parameter or if the system is prompted to give recommendations. Since the system learns from every user response, it can recommend appropriate songs at any point in the test with the information it has learned. In order to do this, the MTT function asks questions based upon the different parameters and decreases the size of the preferred subset according to the songs chosen by the user. Essentially, the system provides the representative clip from two songs previously analyzed and asks the user to select which song the user prefers. By asking binary questions, the system finds the preferred values for each one of the song parameters analyzed.
- As the subset gets smaller, the preferred range for each parameter gets smaller too and, at some point, the user will not be able to distinguish between the two extremes of the parameter for the range found in the subset. In other words, the user will equally like songs from the whole range. The system is able to tell the user's perception level for each parameter and determine that the system cannot learn any more from the user. The MTT function identifies that it has reached this point for a parameter when the user chooses songs indistinctly from both extremes of the subset's range, and so there is no determined consistency in the answers. The MTT function will never have a 100% confidence level for a user (as long as there remain songs in the database); thus, in reality, a new pair of songs can always be requested.
- In theory, the MTT function's learning process seeks to achieve the highest possible confidence level for all parameters, in essence achieving a high confidence level for the user's preferences. In other words, the system seeks to reach a high confidence level for each user and develops a precise taste profile for each user (based on the parameters analyzed). In reality, however, this process can continue endlessly, as the MTT function seeks to shorten preferred ranges and modify preferred values. For commercial applications, it is necessary to set a limit on the MTT process. This limit can be set either by setting a limited number of sound clip comparisons, target confidence level, or both. This is a matter of simple programming on the front-end application on the customer's side.
- A user's profile is stored in the system while the user is still connected to the system. When the session is terminated, the user profile is eliminated. However, this profile can also be saved within the system or external database and accessed at a later date, depending upon the specific retail customer and implementation environment. Thus, a user can go back, continue the MTT process, and obtain refined recommendations. This functionality is achieved using an identification technology such as loyalty cards or cookies, as described in greater detail below.
- The MTT process in a retail situation is adapted to require approximately 8-15 steps to complete the taste test determination process. A progress meter is optionally displayed on the user's screen reflecting the confidence level of the system, thereby allowing users to understand where they are in the test process.
- The typical user interface includes two basic “play buttons” that allow the user to first preview both music clips (multiple times, if they desire), and two basic “select buttons” that then allow the user to select which of the two music clips they prefer. There is no qualitative aspect to these choices; rather, the user simply decides which song is preferred between the two music clips. The music clips used in the MTT process are unique audio files, derived from the clip extraction phase described in detail above.
- The MTT user interface may vary based on implementation environments and operating systems. In fact, the MTT functionality is independent of the user interface as long as the user is presented with the two songs and is able to choose the one that he/she prefers. For example, a user interface could consist of two songs displayed as text and then the user is prompted to type the name of his preferred song, the MTT function will continue to function the same. Furthermore, the interface can be customized to meet the specific needs/requirements of different customers. This includes adding a customer's logo, color scheme, wording/tonality, and audio cues.
- Once the MTT function has been completed, the user's particular musical tastes are established in the user's personal ‘taste’ vector. The system uses the ‘taste’ vector to determine and display a list of song matches to the user's determined musical taste. Alternately, the list of song matches may be displayed to the user at his or her request before completion of the analysis, in which case the recommendations will simply reflect the MTT function's current confidence level. To generate the list of song matches, the MTT function searches the music database in order to find songs that most closely match that user's ‘taste’ profile. Closeness of a song to the ‘taste’ vector is determined by the mathematical distance between two songs.
- The theory of the recommendation engine is based upon the relative correlation between a user's preferred values and each song (where correlation is defined as the relative distance measured as the sum of the squared difference between each parameter). A song is always considered as a whole and, thus, individual parameters are not considered to be matches by themselves. This formula can be illustrated as follows:
where C is the proximity value between the song and the user's preferred values, N is the number of parameters in the parameter vector, S is the parameter vector of the songs in the database, and V is the parameter vector with the user's preferred values. These results can be customized and displayed according to closest match, genre, or release date. Theoretically, only songs with a value of C below a predetermined threshold would be included in the recommendation list. - In order to listen to the recommendations, the user once again presses the appropriate button on their interface device or alternate selection method. Next to each recommendation, there is also a “More Like This” button that displays songs that are close matches to that particular song.
- The “More Like This” (“MLT”) function is implemented through a pattern-recognizing artificial intelligence system using an MLT daemon (labeled ‘mltd’ in
FIGS. 2 and 3 ). The MLT function allows a user to receive music recommendations by selecting a song and requesting songs that are mathematically similar to that song. Within the implementation environment, the user needs to enter the name of a song into the device or simply click on the “More Like This” button to receive recommendations. - When the MLT function is operable, the system performs real-time MLT recommendations by looking for song files that have the most mathematically similar analysis data to the parent song. The MLT lookup process consists of searching for similar songs by checking the relative distances between all parameters for each song. It is the difference between parameters that determine if the songs are similar or not for the user. Therefore, given a list of songs, each song can have a “More Like This” link to similar music.
- This is precisely the same process used to look up recommendations within the MTT function. However, instead of using the user's preferred values, the system looks at the parameter vector for the chosen song.
where C is the proximity value between the song and the user's preferred values, N is the number of parameters in the parameter vector, S is the parameter vector of the songs in the database, and M is the parameter vector with the chosen “More Like This” song. Once again, only songs with a value of C below a predetermined threshold would be included in the “More Like This” list. - In addition to the MLT and MTT functions, a “My Personal Preferences” (“MPP”) function enables a user to establish a personal preference file that may in turn be used by the system to provide the user with music selections that match that user's personal profile. While interacting with the device or website, a user ranks songs against preferably a simple five level scale (e.g., “I love it”, “I like it”, “I neither like nor dislike it”, “I don't like it”, and “I strongly dislike it”) by choosing the appropriate option on a user interface screen. This process can be done individually for each song listened to, or by entering song names and their rankings into the device. The user can add, erase, or edit songs and rankings from their list at any time. When requested, recommendations can be made using the current songs and ratings on the list.
- To do this, the parameter vectors for the MPP selected songs are weighted, based on a user's rankings, and are input to the system. This process is similar to the MTT learning process except, instead of the system deciding what it wants to learn and asking the appropriate questions, the user chooses the songs (thereby choosing the parameter vectors) and forces the system to learn from the user inputs. This is referred to as forced learning.
- Selected songs are grouped based on the given rank in order to find preference groups. The selected songs fall into two types, liked songs and disliked songs. Songs ranked as “I love it”, “I like it”, or “I neither like nor dislike it” are defined as liked songs, and songs ranked as “I don't like it” and “I strongly dislike it” are defined as disliked songs. Similar groupings of songs are important to a successful MPP process.
- Grouping is accomplished by searching through the ranked songs to find commonalities within each type (liked and disliked songs), and then grouping the songs that are most similar (i.e. close in proximity to each other). Note that a person can have more than one preference grouping within each preference group type (e.g. someone can both like slow music and hard rap music). Thus, preference groups are also referred to as musical tastes (and are also either liked or disliked).
- Grouping in this way is also done to eliminate songs that are ranked by the user, but dissimilar to other songs of the same type. Songs are eliminated in this way because they are not valuable to the MPP recommendation process (i.e., the system cannot determine what it is that the user likes from that song).
- Mean values are calculated for each musical taste to determine the corresponding parameter vector (referred to as a musical taste vector). To do this, songs within each musical taste are weighted according to the user's ranking for those songs, thus giving less importance to the “I like it” songs than to the “I love it” songs, but more than to the “I neither like nor dislike it” songs, for the liked musical tastes; and giving more importance to the “I strongly dislike it” than to the “I don't like it”, for the disliked musical tastes. By doing this, the system biases the mean towards the more liked songs and the more disliked songs.
- After the rankings have been established, the system is ready to recommend songs to the user. To do this, the system needs to create a list of possible songs to be recommended. The initial list contains all songs in the database. First, songs that are known to be disliked by the user are erased by analyzing in which relative clusters (within the whole database of songs) the disliked musical taste vectors fall (see discussion above concerning elimination of analysis data correlations). All songs from the clusters containing disliked musical taste vectors are erased from the list of possible songs to be recommended.
- Using the MLT recommendation technique detailed above, recommendations are generated from the liked musical taste vectors using only the list of possible songs to be recommended. The system can either present the musical tastes found to the user (from which to select their recommendations) or simply deliver a combined recommendation list including songs from all different musical tastes.
- In addition to providing such recommendation functions, both the MTT and MLT functions also create user tracking statistics in the form of log files. Such log file information is of a similar format to most daemon software and includes information as discrete calls to the recommendation engine including session number, the device requesting the action, date and time, and the outcome of the action. Every call creates a log entry and thus, unique user sessions within the Human Media Interface system can be tracked and monitored. An example log excerpt is as follows:
- (HMIMData) Time 1031298303 Operation: Calculate Mean (Timer). 0.0000
- (HMIMData) Time 1031298303 Operation: Calculate Standard Deviation (Timer). 0.3200
- (HMIMData) Time 1031298304 Operation: Calculate Sigma (Timer). 0.3600
- (HMIMData) Time 1031298304 Operation: Calculate New Song Data (Timer). 0.1300
- (HMIMData) Time 1031298305 Operation: Calculate Matrix (Timer). 1.2900
- (HMIMData) Time 1031298305 Operation: Calculate Used Parameters (Timer). 0.0000
- (HMIMData) Time 1031298305 Operation: Resort Data (Timer). 0.0200
- (IPServer) HMI TCP/IP Server. Time 1031298310 Connection No.0 New Connection Started from 192.168.0.70
- (IPServer) HMI TCP/IP Server. Time 1031298310 Connection No.0 HELO 8 23
- (IPServer) HMI TCP/IP Server. Time 1031298310 Connection No.0 MLTS “075678326820#1” 20
- (HMIMData) Time 1031298311 Operation: More Like This (Timer). 0.1700
- (IPServer) HMI TCP/IP Server. Time 1031298311 Connection No.0 QUIT
- (IPServer) HMI TCP/IP Server. Time 1031298311 Connection No.0 Connection Closed
- In this example, “HELO 8 23” gives both the third party's id number: 8 (e.g. the retailer) and the terminal id (the actual in-store device that is requesting this transaction: 23).
- Anonymous user session data (not specific to an individual) can be used to determine general trends within the system and therefore is valuable from a business/marketing perspective. Examples of anonymous user session data include number of users per time period, user paths through the system, most/least recommended music, error messages, and average length of time per session.
- User session information can also be attributed to an individual (e.g. John Smith logged into
music kiosk # 2 at 4:30 PM on Aug. 23, 2002). This can be done via the use of scanable loyalty cards or cookies in a web environment. A loyalty card (or customer card) in this case is a card-like physical identifier, such as those currently used by several retail chains and supermarkets. On arrival to a kiosk or in-store music device, the user is prompted to scan their card, thereby enabling the system to associate a specific user session to a person. Online this is done via “cookies”, as is commonly used on the Internet to identify users. - Associated user session data is more robust, as various sessions from the same user can be linked together and personal information garnered. The core of this personal information includes but is not limited to personal music preferences (as deemed by the MTT and MLT functions). This is extremely useful business information for personalized sales and marketing efforts. An example of this includes presenting an individual a personalized music coupon at checkout, offering a CD from which they have already previewed in the system.
- Furthermore, such personalized user session data information can be combined with other marketing databases and exported to various CRM systems, thereby enabling an even higher level of business intelligence. At this level, a retailer can use their unified CRM system to combine an individual's system-derived music preferences, select a CD to be recommended to the individual, check current inventory, determine dynamic pricing for the CD, and deliver a personalized offer via email.
- In an alternate embodiment, the technology of the present invention can be used to predict the potential commercial success of a new song. First, a large database is analyzed to determine a song vector for each song, as described above. In the present embodiment, approximately 3.5 million songs including almost everything that has been released by the music labels since the 1950's until the present time has been analyzed. The database is updated weekly with new releases. Each song is then mapped onto a grid called the music universe and is positioned according to its mathematical characteristics. Each song is represented by a dot on the universe and the songs on one end of the universe are vastly different from songs on the other end of the universe. Songs with mathematical similarities are positioned very close to one another. Next, all of the songs from the music universe that had not been hits in the past five years were removed. The songs remaining were grouped into a limited number of small clusters all over the universe but with vast spaces between them, as illustrated in
FIG. 4 . - “Hit” songs seem to have common characteristics based on the mathematical analysis. While other factors, such as lyrics, theme and artist, influence the success of a song, it is rare for a song that falls outside of the clusters to become a hit.
- In this embodiment, the present invention takes the analyzed song data for the entire database and overlays extra parameters relating to the commercial success of the music in the database. These parameters may include additional data such as total sales, highest chart position, date of release, and other common indicators of commercial success. The specific weight given to such new parameters can be varied to meet an end users specific needs. Using this extra dimension, new releases, potential releases, and even unsigned acts can be compared with the database to see how well a given song fits into the current market and to identify potential hits. As the market changes, the system reflects such changes by finding new patterns in the hit clusters and applying these to the process.
- The system allows for trends to be identified as they develop over time, meaning that a song that contains strong characteristics that are becoming more prevalent in new music and less of the characteristics that are diminishing can be identified as having high potential. Simply put this means that a song that sounds uncommercial to a human listening to it now may just be ahead of its time and in fact contains the right ingredients to appeal to the public sometime in the future.
- For this embodiment, a database of past chart hits from either the US or the UK is used, as well as a database of “classic hits” from 1956 to the present time as the universe of hits. The “hit” database in an archive of every song that has been in the US Weekly Top 30 or the UK Official Charts for the past 5 years. Songs are weighted based upon their ranking within the Top 30. As new songs are added, old songs are deleted. This is done to keep the database as fresh as possible, effectively following the trends and styles as music changes over time.
- A user can submit one, or more, digital music files for analysis. The system scores a new song according to its similarities with current hit songs. First, a parameter vector is determined for the new song. The patterns in such new song are compared to patterns in recent chart “hits” and to patterns in classic hit songs going back to 1956. The new song is compared to the “hits,” by comparing the new song parameter vector to the “hit” parameter vector, as described above, to obtain an affinity value for each song in the “hit” database. The affinity value is a rating that shows how closely related the mathematical patterns in one song are to another. The lower the affinity value between two songs the more closely related they are. An affinity value under 1.00 denotes close similarity while an affinity value of over 1.00 denotes similarity but less obvious and more distant similarity. (When the same recording of a song is compared against itself, the affinity value is zero.) The greater the number of past “hits” with lower affinity values to the new song, the higher the “hit” potential of the new song.
FIG. 5 is an illustration of 12 new songs as analyzed and graphed on the universe of “hits”. In this illustration, there are two songs that fall squarely within hit clusters. - According to the new song's distance with each song in the “hit” song database and the songs within its cluster the system gives the new song a score. By measuring the new song's overall closeness to the “hits” and where such song fits into the “hit” database, an overall score is assigned to the new song. A rating of 7.00 or greater should be considered to be a very strong song and very closely related to what is currently popular in the market. In the same way, using release date and sales information for each song in the “hit” song database the system can give the new song a recentness score and a sales score. A report is provided to the user who submitted the new song indicating the overall score, the proximity to “hit” songs, and the proximity to current “hit” songs. Using such information, the user can determine an appropriate course of action for commercializing the analyzed new song.
- The invention has been described with references to a preferred embodiment. While specific values, relationships, materials and steps have been set forth for purposes of describing concepts of the invention, it will be appreciated by persons skilled in the art that numerous variations and/or modifications may be made to the invention as shown in the specific embodiments without departing from the spirit or scope of the basic concepts and operating principles of the invention as broadly described. It should be recognized that, in the light of the above teachings, those skilled in the art can modify those specifics without departing from the invention taught herein. Having now fully set forth the preferred embodiments and certain modifications of the concept underlying the present invention, various other embodiments as well as certain variations and modifications of the embodiments herein shown and described will obviously occur to those skilled in the art upon becoming familiar with such underlying concept. It is intended to include all such modifications, alternatives and other embodiments insofar as they come within the scope of the appended claims or equivalents thereof. It should be understood, therefore, that the invention may be practiced otherwise than as specifically set forth herein. Consequently, the present embodiments are to be considered in all respects as illustrative and not restrictive.
Claims (8)
1. A method of analyzing music, said method comprising the steps of:
a) providing a digital database comprising a plurality of digital song files;
b) selecting one of said song files for analysis;
c) dividing said selected song file into a plurality of discrete parts;
d) using Fast Fourier Transform techniques on each part of said selected song file to establish a plurality of coefficients, wherein said coefficients are representative of predetermined quantifiable characteristics of said selected song, and;
wherein each said predetermined characteristic is a physical parameter based on human perception selected from the group consisting of:
brightness;
bandwidth;
tempo;
volume;
rhythm;
low frequency;
noise; and
octave, and
how said parameters change over time;
e) determining an average value of the coefficients for each characteristic from each said part of said selected song file;
f) compiling a song vector comprising a sequential list of said average values of the coefficients for each said characteristic for said selected song file; and
g) repeating steps b) through f) for each song in said database.
2. The method according to claim 1 , wherein said digital database comprises a plurality of compressed digital song files, said method further comprising the step of:
b1) decompressing said selected song file prior to dividing said selected song file into a plurality of discrete parts.
3. A method of determining a user's music preference, said method comprising the steps of:
a) providing a digital database comprising a plurality of digital song files;
b) mathematically analyzing each said digital song file to determine a numerical value for a plurality of selected quantifiable characteristics;
wherein said characteristic is a physical parameter based on human perception comprising:
brightness;
bandwidth;
tempo;
volume;
rhythm;
low frequency;
noise; and
octave, and
how said parameters change over time;
c) compiling a song vector comprising a sequential list of said numerical values for each of said plurality of selected characteristic for each said song file;
d) dividing each said song file into portions of selected size and mathematically analyzing each said portion to determine a numerical value for said plurality of selected characteristics for each said portion and compiling a portion vector comprising a sequential list of numerical values for each of said plurality of characteristics for each said portion;
e) selecting and storing a representative portion of each said song file wherein the portion vector of said representative portion substantially mathematically matches the song vector of said song file;
f) choosing two dissimilar representative portions and enabling said user to listen to both representative portions;
g) permitting said user to indicate which of said two dissimilar representative portions said user prefers; and
h) repeating steps f) and g), as necessary, to establish a taste vector for said user comprising song characteristics that said user prefers.
4. The method according to claim 3 , said mathematically analyzing steps further comprising the step of:
using fast Fourier Transform techniques to establish a plurality of coefficients, wherein said coefficients are representative of said characteristics of said song.
5. The method according to claim 3 , wherein said method is performed via a computer website.
6. A method of comparing a new song to previously commercially successful songs, said method comprising:
a) establishing a digital database comprising a plurality of digital song files wherein said songs have been identified as commercially successful;
b) mathematically analyzing each said digital song file to determine a numerical value for a plurality of selected quantifiable characteristics;
wherein said characteristic is a physical parameter based on human perception selected from the group consisting of:
brightness;
bandwidth;
tempo;
volume;
rhythm;
low frequency;
noise; and
octave, and
how said parameters change over time;
c) compiling a song vector comprising a sequential list of said numerical values for each of said plurality of selected characteristic for each said song file;
d) presenting said new song as a digital music file for comparison;
e) mathematically analyzing said new song file to determine a numerical value for the same plurality of selected quantifiable characteristics;
f) compiling a new song vector comprising a sequential list of said numerical values for each of said plurality of selected characteristic for said new song file;
g) establishing an affinity value for said new song as compared to each song vector in the database by summing the square of the difference between the numerical values of each characteristic in each said vector; and
j) determining the potential for commercial success if said affinity value is below a predetermined threshold.
7. The method according to claim 6 , said mathematically analyzing steps further comprising the step of:
using fast Fourier Transform techniques to establish a plurality of coefficients, wherein said coefficients are representative of said characteristics of said song.
8. The method according to claim 6 , wherein said new song is presented via a computer network.
Priority Applications (7)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US11/492,355 US20060254411A1 (en) | 2002-10-03 | 2006-07-25 | Method and system for music recommendation |
US11/825,457 US8053659B2 (en) | 2002-10-03 | 2007-07-06 | Music intelligence universe server |
US11/881,810 US8063295B2 (en) | 2002-10-03 | 2007-07-27 | Method and system for video and film recommendation |
US12/753,169 US7982117B2 (en) | 2002-10-03 | 2010-04-02 | Music intelligence universe server |
US13/174,904 US8338685B2 (en) | 2002-10-03 | 2011-07-01 | Method and system for video and film recommendation |
US13/291,227 US8519249B2 (en) | 2002-10-03 | 2011-11-08 | Music intelligence universe server |
US13/726,874 US8841534B2 (en) | 2002-10-03 | 2012-12-26 | Method and system for video and film recommendation |
Applications Claiming Priority (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US41586802P | 2002-10-03 | 2002-10-03 | |
US10/678,505 US7081579B2 (en) | 2002-10-03 | 2003-10-03 | Method and system for music recommendation |
US11/492,355 US20060254411A1 (en) | 2002-10-03 | 2006-07-25 | Method and system for music recommendation |
Related Parent Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US10/678,505 Continuation US7081579B2 (en) | 2002-10-03 | 2003-10-03 | Method and system for music recommendation |
Related Child Applications (2)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US11/825,457 Continuation-In-Part US8053659B2 (en) | 2002-10-03 | 2007-07-06 | Music intelligence universe server |
US11/881,810 Continuation-In-Part US8063295B2 (en) | 2002-10-03 | 2007-07-27 | Method and system for video and film recommendation |
Publications (1)
Publication Number | Publication Date |
---|---|
US20060254411A1 true US20060254411A1 (en) | 2006-11-16 |
Family
ID=32474411
Family Applications (2)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US10/678,505 Expired - Lifetime US7081579B2 (en) | 2002-10-03 | 2003-10-03 | Method and system for music recommendation |
US11/492,355 Abandoned US20060254411A1 (en) | 2002-10-03 | 2006-07-25 | Method and system for music recommendation |
Family Applications Before (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US10/678,505 Expired - Lifetime US7081579B2 (en) | 2002-10-03 | 2003-10-03 | Method and system for music recommendation |
Country Status (1)
Country | Link |
---|---|
US (2) | US7081579B2 (en) |
Cited By (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20050034150A1 (en) * | 2003-08-07 | 2005-02-10 | Sony Corporation | Server, content providing apparatus, content receiving apparatus, content providing method, content receiving method, and program |
US20070157797A1 (en) * | 2005-12-14 | 2007-07-12 | Sony Corporation | Taste profile production apparatus, taste profile production method and profile production program |
US20070219996A1 (en) * | 2006-03-17 | 2007-09-20 | Vervelife | System and method for creating custom playlists based on user inputs |
US20080228744A1 (en) * | 2007-03-12 | 2008-09-18 | Desbiens Jocelyn | Method and a system for automatic evaluation of digital files |
US20090182736A1 (en) * | 2008-01-16 | 2009-07-16 | Kausik Ghatak | Mood based music recommendation method and system |
US20090228796A1 (en) * | 2008-03-05 | 2009-09-10 | Sony Corporation | Method and device for personalizing a multimedia application |
US20090281906A1 (en) * | 2008-05-07 | 2009-11-12 | Microsoft Corporation | Music Recommendation using Emotional Allocation Modeling |
US20090277322A1 (en) * | 2008-05-07 | 2009-11-12 | Microsoft Corporation | Scalable Music Recommendation by Search |
US20100057734A1 (en) * | 2008-09-03 | 2010-03-04 | Yasushi Miyajima | Music processing method, music processing apparatus and program |
US20100300271A1 (en) * | 2009-05-27 | 2010-12-02 | Microsoft Corporation | Detecting Beat Information Using a Diverse Set of Correlations |
US20100318544A1 (en) * | 2009-06-15 | 2010-12-16 | Telefonaktiebolaget Lm Ericsson (Publ) | Device and method for selecting at least one media for recommendation to a user |
US20120310925A1 (en) * | 2011-06-06 | 2012-12-06 | Dmitry Kozko | System and method for determining art preferences of people |
US20140143222A1 (en) * | 2012-11-16 | 2014-05-22 | Google Inc. | Ranking signals for sparse corpora |
TWI458310B (en) * | 2007-08-22 | 2014-10-21 | Microsoft Corp | Collaborative media recommendatin and sharing technique |
TWI584139B (en) * | 2016-09-09 | 2017-05-21 | Speech recognition system and its information processing method applied to non - real - time signal source | |
US20190325894A1 (en) * | 2014-06-30 | 2019-10-24 | Dolby Laboratories Licensing Corporation | Perception based multimedia processing |
WO2023232865A1 (en) | 2022-06-03 | 2023-12-07 | Xandrie SA | Electronic system for digital and physical media |
Families Citing this family (160)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20020002039A1 (en) | 1998-06-12 | 2002-01-03 | Safi Qureshey | Network-enabled audio device |
US20050038819A1 (en) * | 2000-04-21 | 2005-02-17 | Hicken Wendell T. | Music Recommendation system and method |
US7013301B2 (en) * | 2003-09-23 | 2006-03-14 | Predixis Corporation | Audio fingerprinting system and method |
BR0015441A (en) * | 1999-11-10 | 2003-07-15 | Launch Media Inc | Internet radio and broadcasting method |
US6389467B1 (en) | 2000-01-24 | 2002-05-14 | Friskit, Inc. | Streaming media search and continuous playback system of media resources located by multiple network addresses |
US20060217828A1 (en) * | 2002-10-23 | 2006-09-28 | Hicken Wendell T | Music searching system and method |
AU2001271980B2 (en) * | 2000-07-11 | 2004-07-29 | Excalibur Ip, Llc | Online playback system with community bias |
US7918738B2 (en) * | 2001-03-27 | 2011-04-05 | Igt | Interactive game playing preferences |
US8480466B2 (en) * | 2001-03-27 | 2013-07-09 | Igt | Method and apparatus for previewing a game |
CA2479838C (en) * | 2002-03-19 | 2011-02-08 | Mapinfo Corporation | Location based service provider |
US9172915B2 (en) * | 2004-08-04 | 2015-10-27 | Dizpersion Corporation | Method of operating a channel recommendation system |
DE10232916B4 (en) * | 2002-07-19 | 2008-08-07 | Fraunhofer-Gesellschaft zur Förderung der angewandten Forschung e.V. | Apparatus and method for characterizing an information signal |
US20070261077A1 (en) * | 2006-05-08 | 2007-11-08 | Gary Zalewski | Using audio/visual environment to select ads on game platform |
US20070061413A1 (en) * | 2005-09-15 | 2007-03-15 | Larsen Eric J | System and method for obtaining user information from voices |
US20070260517A1 (en) * | 2006-05-08 | 2007-11-08 | Gary Zalewski | Profile detection |
US7083520B2 (en) * | 2002-09-11 | 2006-08-01 | Igt | In-room game promotion and demonstration method and system |
US8063295B2 (en) * | 2002-10-03 | 2011-11-22 | Polyphonic Human Media Interface, S.L. | Method and system for video and film recommendation |
US8053659B2 (en) * | 2002-10-03 | 2011-11-08 | Polyphonic Human Media Interface, S.L. | Music intelligence universe server |
US8686272B2 (en) * | 2002-10-03 | 2014-04-01 | Polyphonic Human Media Interface, S.L. | Method and system for music recommendation based on immunology |
US7723602B2 (en) * | 2003-08-20 | 2010-05-25 | David Joseph Beckford | System, computer program and method for quantifying and analyzing musical intellectual property |
JP2005301921A (en) * | 2004-04-15 | 2005-10-27 | Sharp Corp | Musical composition retrieval system and musical composition retrieval method |
US8028323B2 (en) | 2004-05-05 | 2011-09-27 | Dryden Enterprises, Llc | Method and system for employing a first device to direct a networked audio device to obtain a media item |
US9826046B2 (en) | 2004-05-05 | 2017-11-21 | Black Hills Media, Llc | Device discovery for digital entertainment network |
US8028038B2 (en) | 2004-05-05 | 2011-09-27 | Dryden Enterprises, Llc | Obtaining a playlist based on user profile matching |
US20070276733A1 (en) * | 2004-06-23 | 2007-11-29 | Frank Geshwind | Method and system for music information retrieval |
US20060212149A1 (en) * | 2004-08-13 | 2006-09-21 | Hicken Wendell T | Distributed system and method for intelligent data analysis |
JP4447989B2 (en) * | 2004-08-27 | 2010-04-07 | ソニー株式会社 | Data reproducing apparatus, data reproducing method and data processing system |
US20060083119A1 (en) * | 2004-10-20 | 2006-04-20 | Hayes Thomas J | Scalable system and method for predicting hit music preferences for an individual |
US7777125B2 (en) * | 2004-11-19 | 2010-08-17 | Microsoft Corporation | Constructing a table of music similarity vectors from a music similarity graph |
JP4655200B2 (en) * | 2005-02-01 | 2011-03-23 | ソニー株式会社 | Information processing apparatus and method, and program |
US7734569B2 (en) * | 2005-02-03 | 2010-06-08 | Strands, Inc. | Recommender system for identifying a new set of media items responsive to an input set of media items and knowledge base metrics |
CA2644943A1 (en) * | 2005-03-04 | 2006-09-14 | Musicip Corporation | Scan shuffle for building playlists |
JP4816207B2 (en) | 2005-04-01 | 2011-11-16 | ソニー株式会社 | Information processing system and method |
KR100731761B1 (en) * | 2005-05-02 | 2007-06-22 | 주식회사 싸일런트뮤직밴드 | Music production system and method by using internet |
US7613736B2 (en) * | 2005-05-23 | 2009-11-03 | Resonance Media Services, Inc. | Sharing music essence in a recommendation system |
US8645985B2 (en) * | 2005-09-15 | 2014-02-04 | Sony Computer Entertainment Inc. | System and method for detecting user attention |
US8616973B2 (en) * | 2005-09-15 | 2013-12-31 | Sony Computer Entertainment Inc. | System and method for control by audible device |
EP1952281A2 (en) * | 2005-11-16 | 2008-08-06 | Koninklijke Philips Electronics N.V. | Method of generating and methods of filtering a user profile |
JP2007172138A (en) * | 2005-12-20 | 2007-07-05 | Sony Corp | Content reproduction device, list correction unit, content reproduction method and list correction method |
KR100749045B1 (en) * | 2006-01-26 | 2007-08-13 | 삼성전자주식회사 | Method and apparatus for searching similar music using summary of music content |
KR100717387B1 (en) * | 2006-01-26 | 2007-05-11 | 삼성전자주식회사 | Method and apparatus for searching similar music |
US8510453B2 (en) * | 2007-03-21 | 2013-08-13 | Samsung Electronics Co., Ltd. | Framework for correlating content on a local network with information on an external network |
US8843467B2 (en) * | 2007-05-15 | 2014-09-23 | Samsung Electronics Co., Ltd. | Method and system for providing relevant information to a user of a device in a local network |
US8285595B2 (en) * | 2006-03-29 | 2012-10-09 | Napo Enterprises, Llc | System and method for refining media recommendations |
US20070244751A1 (en) * | 2006-04-17 | 2007-10-18 | Gary Zalewski | Using visual environment to select ads on game platform |
US20070243930A1 (en) * | 2006-04-12 | 2007-10-18 | Gary Zalewski | System and method for using user's audio environment to select advertising |
US20070255630A1 (en) * | 2006-04-17 | 2007-11-01 | Gary Zalewski | System and method for using user's visual environment to select advertising |
US7772478B2 (en) * | 2006-04-12 | 2010-08-10 | Massachusetts Institute Of Technology | Understanding music |
US7603434B2 (en) * | 2006-04-13 | 2009-10-13 | Domingo Enterprises, Llc | Central system providing previews of a user's media collection to a portable media player |
US20070245376A1 (en) * | 2006-04-13 | 2007-10-18 | Concert Technology Corporation | Portable media player enabled to obtain previews of media content |
US20070244985A1 (en) * | 2006-04-13 | 2007-10-18 | Concert Technology Corporation | User system providing previews of a user's media collection to an associated portable media player |
US20070245378A1 (en) * | 2006-04-13 | 2007-10-18 | Concert Technology Corporation | User system providing previews to an associated portable media player |
US8316081B2 (en) | 2006-04-13 | 2012-11-20 | Domingo Enterprises, Llc | Portable media player enabled to obtain previews of a user's media collection |
US20070245377A1 (en) * | 2006-04-13 | 2007-10-18 | Concert Technology Corporation | Central system providing previews to a portable media player |
US20070282860A1 (en) * | 2006-05-12 | 2007-12-06 | Marios Athineos | Method and system for music information retrieval |
US8903843B2 (en) | 2006-06-21 | 2014-12-02 | Napo Enterprises, Llc | Historical media recommendation service |
JP2008015595A (en) * | 2006-07-03 | 2008-01-24 | Sony Corp | Content selection recommendation method, server, content reproduction device, content recording device and program for selecting and recommending of content |
US9003056B2 (en) | 2006-07-11 | 2015-04-07 | Napo Enterprises, Llc | Maintaining a minimum level of real time media recommendations in the absence of online friends |
US8805831B2 (en) * | 2006-07-11 | 2014-08-12 | Napo Enterprises, Llc | Scoring and replaying media items |
US8059646B2 (en) * | 2006-07-11 | 2011-11-15 | Napo Enterprises, Llc | System and method for identifying music content in a P2P real time recommendation network |
US8327266B2 (en) | 2006-07-11 | 2012-12-04 | Napo Enterprises, Llc | Graphical user interface system for allowing management of a media item playlist based on a preference scoring system |
US7970922B2 (en) | 2006-07-11 | 2011-06-28 | Napo Enterprises, Llc | P2P real time media recommendations |
US7680959B2 (en) * | 2006-07-11 | 2010-03-16 | Napo Enterprises, Llc | P2P network for providing real time media recommendations |
US8620699B2 (en) * | 2006-08-08 | 2013-12-31 | Napo Enterprises, Llc | Heavy influencer media recommendations |
US8090606B2 (en) | 2006-08-08 | 2012-01-03 | Napo Enterprises, Llc | Embedded media recommendations |
US7645929B2 (en) * | 2006-09-11 | 2010-01-12 | Hewlett-Packard Development Company, L.P. | Computational music-tempo estimation |
US9256602B2 (en) * | 2006-09-13 | 2016-02-09 | Cellco Partnership | System and method for distributing and providing recommendations related to playable content to a user based on information extracted from one or more playback devices of the user |
US9008634B2 (en) * | 2006-10-06 | 2015-04-14 | Napo Enterprises, Llc | System and method for providing media content selections |
US8935269B2 (en) * | 2006-12-04 | 2015-01-13 | Samsung Electronics Co., Ltd. | Method and apparatus for contextual search and query refinement on consumer electronics devices |
US8874655B2 (en) * | 2006-12-13 | 2014-10-28 | Napo Enterprises, Llc | Matching participants in a P2P recommendation network loosely coupled to a subscription service |
US20090070185A1 (en) * | 2007-01-17 | 2009-03-12 | Concert Technology Corporation | System and method for recommending a digital media subscription service |
WO2008100431A1 (en) | 2007-02-09 | 2008-08-21 | Cyberscan Technology Inc. | Universal player control for casino game graphic assets |
JP4730619B2 (en) * | 2007-03-02 | 2011-07-20 | ソニー株式会社 | Information processing apparatus and method, and program |
US20080222088A1 (en) * | 2007-03-08 | 2008-09-11 | Motorola, Inc. | Apparatus and Methods for Generating an Implicit Vote for a Media Item Based on User Behavior Relative to a Media Client |
US20080218814A1 (en) * | 2007-03-09 | 2008-09-11 | Andrew Rodney Ferlitsch | Color Proofing Method |
US9224427B2 (en) * | 2007-04-02 | 2015-12-29 | Napo Enterprises LLC | Rating media item recommendations using recommendation paths and/or media item usage |
US8112720B2 (en) | 2007-04-05 | 2012-02-07 | Napo Enterprises, Llc | System and method for automatically and graphically associating programmatically-generated media item recommendations related to a user's socially recommended media items |
US7949649B2 (en) * | 2007-04-10 | 2011-05-24 | The Echo Nest Corporation | Automatically acquiring acoustic and cultural information about music |
US8073854B2 (en) * | 2007-04-10 | 2011-12-06 | The Echo Nest Corporation | Determining the similarity of music using cultural and acoustic information |
US7985911B2 (en) | 2007-04-18 | 2011-07-26 | Oppenheimer Harold B | Method and apparatus for generating and updating a pre-categorized song database from which consumers may select and then download desired playlists |
US20080257134A1 (en) * | 2007-04-18 | 2008-10-23 | 3B Music, Llc | Method And Apparatus For Generating And Updating A Pre-Categorized Song Database From Which Consumers May Select And Then Download Desired Playlists |
US8285776B2 (en) * | 2007-06-01 | 2012-10-09 | Napo Enterprises, Llc | System and method for processing a received media item recommendation message comprising recommender presence information |
US9164993B2 (en) * | 2007-06-01 | 2015-10-20 | Napo Enterprises, Llc | System and method for propagating a media item recommendation message comprising recommender presence information |
US20080301187A1 (en) * | 2007-06-01 | 2008-12-04 | Concert Technology Corporation | Enhanced media item playlist comprising presence information |
US9037632B2 (en) * | 2007-06-01 | 2015-05-19 | Napo Enterprises, Llc | System and method of generating a media item recommendation message with recommender presence information |
US8839141B2 (en) | 2007-06-01 | 2014-09-16 | Napo Enterprises, Llc | Method and system for visually indicating a replay status of media items on a media device |
US20090049045A1 (en) | 2007-06-01 | 2009-02-19 | Concert Technology Corporation | Method and system for sorting media items in a playlist on a media device |
US20090006397A1 (en) * | 2007-06-29 | 2009-01-01 | Nokia Corporation | Systems, Methods, Devices and Computer Program Products for Providing Media Recommendation Trekking |
WO2009074871A2 (en) * | 2007-07-05 | 2009-06-18 | Avet Manukyan | System and method for the characterization, selection and recommendation of digital music and media content |
US20090049030A1 (en) * | 2007-08-13 | 2009-02-19 | Concert Technology Corporation | System and method for reducing the multiple listing of a media item in a playlist |
US20090048992A1 (en) * | 2007-08-13 | 2009-02-19 | Concert Technology Corporation | System and method for reducing the repetitive reception of a media item recommendation |
US8583615B2 (en) * | 2007-08-31 | 2013-11-12 | Yahoo! Inc. | System and method for generating a playlist from a mood gradient |
US7865522B2 (en) * | 2007-11-07 | 2011-01-04 | Napo Enterprises, Llc | System and method for hyping media recommendations in a media recommendation system |
US9060034B2 (en) | 2007-11-09 | 2015-06-16 | Napo Enterprises, Llc | System and method of filtering recommenders in a media item recommendation system |
US20090144273A1 (en) * | 2007-12-03 | 2009-06-04 | Paul Kappos | System and method for music and compatibility matching |
US9224150B2 (en) | 2007-12-18 | 2015-12-29 | Napo Enterprises, Llc | Identifying highly valued recommendations of users in a media recommendation network |
US8396951B2 (en) * | 2007-12-20 | 2013-03-12 | Napo Enterprises, Llc | Method and system for populating a content repository for an internet radio service based on a recommendation network |
US9734507B2 (en) * | 2007-12-20 | 2017-08-15 | Napo Enterprise, Llc | Method and system for simulating recommendations in a social network for an offline user |
US8060525B2 (en) * | 2007-12-21 | 2011-11-15 | Napo Enterprises, Llc | Method and system for generating media recommendations in a distributed environment based on tagging play history information with location information |
US8316015B2 (en) | 2007-12-21 | 2012-11-20 | Lemi Technology, Llc | Tunersphere |
US8117193B2 (en) | 2007-12-21 | 2012-02-14 | Lemi Technology, Llc | Tunersphere |
JP2011511384A (en) * | 2008-02-07 | 2011-04-07 | ブランド・アフィニティー・テクノロジーズ・インコーポレイテッド | Qualitative and quantitative methods for rating brands using keywords |
US8642872B2 (en) * | 2008-03-03 | 2014-02-04 | Microsoft Corporation | Music steering with automatically detected musical attributes |
US8725740B2 (en) | 2008-03-24 | 2014-05-13 | Napo Enterprises, Llc | Active playlist having dynamic media item groups |
US20090259621A1 (en) * | 2008-04-11 | 2009-10-15 | Concert Technology Corporation | Providing expected desirability information prior to sending a recommendation |
US8484311B2 (en) | 2008-04-17 | 2013-07-09 | Eloy Technology, Llc | Pruning an aggregate media collection |
US7958130B2 (en) * | 2008-05-26 | 2011-06-07 | Microsoft Corporation | Similarity-based content sampling and relevance feedback |
US20090307053A1 (en) * | 2008-06-06 | 2009-12-10 | Ryan Steelberg | Apparatus, system and method for a brand affinity engine using positive and negative mentions |
EP2159720A1 (en) * | 2008-08-28 | 2010-03-03 | Bach Technology AS | Apparatus and method for generating a collection profile and for communicating based on the collection profile |
US8938465B2 (en) * | 2008-09-10 | 2015-01-20 | Samsung Electronics Co., Ltd. | Method and system for utilizing packaged content sources to identify and provide information based on contextual information |
US8484227B2 (en) | 2008-10-15 | 2013-07-09 | Eloy Technology, Llc | Caching and synching process for a media sharing system |
US8880599B2 (en) | 2008-10-15 | 2014-11-04 | Eloy Technology, Llc | Collection digest for a media sharing system |
US7994410B2 (en) * | 2008-10-22 | 2011-08-09 | Classical Archives, LLC | Music recording comparison engine |
US8200674B2 (en) * | 2009-01-19 | 2012-06-12 | Microsoft Corporation | Personalized media recommendation |
US8200602B2 (en) | 2009-02-02 | 2012-06-12 | Napo Enterprises, Llc | System and method for creating thematic listening experiences in a networked peer media recommendation environment |
US20100205222A1 (en) * | 2009-02-10 | 2010-08-12 | Tom Gajdos | Music profiling |
US8168876B2 (en) * | 2009-04-10 | 2012-05-01 | Cyberlink Corp. | Method of displaying music information in multimedia playback and related electronic device |
US9460092B2 (en) * | 2009-06-16 | 2016-10-04 | Rovi Technologies Corporation | Media asset recommendation service |
TWI410811B (en) * | 2009-09-08 | 2013-10-01 | Univ Nat Cheng Kung | Music recommendation method and program product thereof |
US8429109B2 (en) | 2009-09-18 | 2013-04-23 | Apple Inc. | Segmented graphical representations for recommending elements |
US9053695B2 (en) * | 2010-03-04 | 2015-06-09 | Avid Technology, Inc. | Identifying musical elements with similar rhythms |
US8400876B2 (en) * | 2010-09-30 | 2013-03-19 | Mitsubishi Electric Research Laboratories, Inc. | Method and system for sensing objects in a scene using transducer arrays and coherent wideband ultrasound pulses |
US8751527B1 (en) | 2010-10-11 | 2014-06-10 | Paul Riggio | Information retrieval system |
WO2012068705A1 (en) * | 2010-11-25 | 2012-05-31 | Telefonaktiebolaget L M Ericsson (Publ) | Analysis system and method for audio data |
CN102004785B (en) * | 2010-11-30 | 2012-08-29 | 北京瑞信在线系统技术有限公司 | Method and device for recommending music |
US20120284069A1 (en) * | 2011-05-04 | 2012-11-08 | Sony Corporation | Method for optimizing parameters in a recommendation system |
GB201109731D0 (en) | 2011-06-10 | 2011-07-27 | System Ltd X | Method and system for analysing audio tracks |
JP2013025555A (en) * | 2011-07-21 | 2013-02-04 | Sony Corp | Information processor, information processing system, information processing method, and, program |
EP2549423A1 (en) * | 2011-07-22 | 2013-01-23 | Axel Springer Digital TV Guide GmbH | Automatic determination of the relevance of recommendations in a social network |
US10262327B1 (en) * | 2011-09-22 | 2019-04-16 | Glance Networks, Inc. | Integrating screen sharing sessions with customer relationship management |
US8671068B2 (en) | 2011-09-22 | 2014-03-11 | Toyota Jidosha Kabushiki Kaisha | Content recommendation system |
US8909667B2 (en) | 2011-11-01 | 2014-12-09 | Lemi Technology, Llc | Systems, methods, and computer readable media for generating recommendations in a media recommendation system |
US8492633B2 (en) | 2011-12-02 | 2013-07-23 | The Echo Nest Corporation | Musical fingerprinting |
US8586847B2 (en) * | 2011-12-02 | 2013-11-19 | The Echo Nest Corporation | Musical fingerprinting based on onset intervals |
US9436929B2 (en) * | 2012-01-24 | 2016-09-06 | Verizon Patent And Licensing Inc. | Collaborative event playlist systems and methods |
US20130339853A1 (en) * | 2012-06-18 | 2013-12-19 | Ian Paul Hierons | Systems and Method to Facilitate Media Search Based on Acoustic Attributes |
US9020923B2 (en) | 2012-06-18 | 2015-04-28 | Score Revolution, Llc | Systems and methods to facilitate media search |
US10140372B2 (en) | 2012-09-12 | 2018-11-27 | Gracenote, Inc. | User profile based on clustering tiered descriptors |
CN103970802B (en) * | 2013-02-05 | 2018-12-14 | 北京音之邦文化科技有限公司 | A kind of method and device of song recommendations |
US9383965B1 (en) * | 2013-08-16 | 2016-07-05 | Amazon Technologies, Inc. | Media library analyzer |
US20150348164A1 (en) * | 2014-04-03 | 2015-12-03 | Uma Satish Doshi | Method and system for music recommendation |
US10321842B2 (en) | 2014-04-22 | 2019-06-18 | Interaxon Inc. | System and method for associating music with brain-state data |
US10346754B2 (en) | 2014-09-18 | 2019-07-09 | Sounds Like Me Limited | Method and system for psychological evaluation based on music preferences |
US9736204B2 (en) * | 2015-06-24 | 2017-08-15 | Pandora Media, Inc. | Media content delivery over telephone networks |
US10546016B1 (en) | 2015-11-06 | 2020-01-28 | Gracenote, Inc. | Audiovisual content curation system |
US10832304B2 (en) | 2016-01-15 | 2020-11-10 | Target Brands, Inc. | Resorting product suggestions for a user interface |
JP7100422B2 (en) | 2016-10-21 | 2022-07-13 | 富士通株式会社 | Devices, programs, and methods for recognizing data properties |
ES2765415T3 (en) | 2016-10-21 | 2020-06-09 | Fujitsu Ltd | Microservices-based data processing apparatus, method and program |
EP3312722A1 (en) | 2016-10-21 | 2018-04-25 | Fujitsu Limited | Data processing apparatus, method, and program |
US10776170B2 (en) | 2016-10-21 | 2020-09-15 | Fujitsu Limited | Software service execution apparatus, system, and method |
JP6805765B2 (en) | 2016-10-21 | 2020-12-23 | 富士通株式会社 | Systems, methods, and programs for running software services |
US9934785B1 (en) | 2016-11-30 | 2018-04-03 | Spotify Ab | Identification of taste attributes from an audio signal |
CN108574771A (en) * | 2017-03-10 | 2018-09-25 | 峰范(北京)科技有限公司 | Collecting and processing of information system and its voice playing device, processing method |
GB201718894D0 (en) | 2017-11-15 | 2017-12-27 | X-System Ltd | Russel space |
KR102495888B1 (en) * | 2018-12-04 | 2023-02-03 | 삼성전자주식회사 | Electronic device for outputting sound and operating method thereof |
CN110457492A (en) * | 2019-06-06 | 2019-11-15 | 北京酷我科技有限公司 | A kind of individuation data proposed algorithm and recommender system |
US11188605B2 (en) | 2019-07-31 | 2021-11-30 | Rovi Guides, Inc. | Systems and methods for recommending collaborative content |
CN110781348B (en) * | 2019-10-25 | 2023-06-06 | 北京威晟艾德尔科技有限公司 | Video file analysis method |
US10755291B1 (en) | 2019-10-25 | 2020-08-25 | Isolation Network, Inc. | Artificial intelligence automation of marketing campaigns |
CN113569086B (en) * | 2021-08-05 | 2024-01-26 | 深圳墨世科技有限公司 | Method, device, terminal equipment and readable storage medium for aggregating curved libraries |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5918223A (en) * | 1996-07-22 | 1999-06-29 | Muscle Fish | Method and article of manufacture for content-based analysis, storage, retrieval, and segmentation of audio information |
US20020147628A1 (en) * | 2001-02-16 | 2002-10-10 | Jeffrey Specter | Method and apparatus for generating recommendations for consumer preference items |
US20030045953A1 (en) * | 2001-08-21 | 2003-03-06 | Microsoft Corporation | System and methods for providing automatic classification of media entities according to sonic properties |
Family Cites Families (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5616876A (en) | 1995-04-19 | 1997-04-01 | Microsoft Corporation | System and methods for selecting music on the basis of subjective content |
US5792971A (en) | 1995-09-29 | 1998-08-11 | Opcode Systems, Inc. | Method and system for editing digital audio information with music-like parameters |
JP3601371B2 (en) | 1999-09-27 | 2004-12-15 | ヤマハ株式会社 | Waveform generation method and apparatus |
US20020002899A1 (en) | 2000-03-22 | 2002-01-10 | Gjerdingen Robert O. | System for content based music searching |
AU2003225052A1 (en) * | 2002-04-16 | 2003-11-03 | Sky Kruse | Method and system for watermarking digital content and for introducing failure points into digital content |
-
2003
- 2003-10-03 US US10/678,505 patent/US7081579B2/en not_active Expired - Lifetime
-
2006
- 2006-07-25 US US11/492,355 patent/US20060254411A1/en not_active Abandoned
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5918223A (en) * | 1996-07-22 | 1999-06-29 | Muscle Fish | Method and article of manufacture for content-based analysis, storage, retrieval, and segmentation of audio information |
US20020147628A1 (en) * | 2001-02-16 | 2002-10-10 | Jeffrey Specter | Method and apparatus for generating recommendations for consumer preference items |
US20030045953A1 (en) * | 2001-08-21 | 2003-03-06 | Microsoft Corporation | System and methods for providing automatic classification of media entities according to sonic properties |
Cited By (30)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8640178B2 (en) * | 2003-08-07 | 2014-01-28 | Sony Corporation | Server, content providing apparatus, content receiving apparatus, content providing method, content receiving method, and program |
US20050034150A1 (en) * | 2003-08-07 | 2005-02-10 | Sony Corporation | Server, content providing apparatus, content receiving apparatus, content providing method, content receiving method, and program |
US20070157797A1 (en) * | 2005-12-14 | 2007-07-12 | Sony Corporation | Taste profile production apparatus, taste profile production method and profile production program |
US20070219996A1 (en) * | 2006-03-17 | 2007-09-20 | Vervelife | System and method for creating custom playlists based on user inputs |
US7873634B2 (en) | 2007-03-12 | 2011-01-18 | Hitlab Ulc. | Method and a system for automatic evaluation of digital files |
US20080228744A1 (en) * | 2007-03-12 | 2008-09-18 | Desbiens Jocelyn | Method and a system for automatic evaluation of digital files |
TWI458310B (en) * | 2007-08-22 | 2014-10-21 | Microsoft Corp | Collaborative media recommendatin and sharing technique |
US20090182736A1 (en) * | 2008-01-16 | 2009-07-16 | Kausik Ghatak | Mood based music recommendation method and system |
US8260778B2 (en) | 2008-01-16 | 2012-09-04 | Kausik Ghatak | Mood based music recommendation method and system |
US20090228796A1 (en) * | 2008-03-05 | 2009-09-10 | Sony Corporation | Method and device for personalizing a multimedia application |
US9491256B2 (en) * | 2008-03-05 | 2016-11-08 | Sony Corporation | Method and device for personalizing a multimedia application |
US8344233B2 (en) | 2008-05-07 | 2013-01-01 | Microsoft Corporation | Scalable music recommendation by search |
US8438168B2 (en) | 2008-05-07 | 2013-05-07 | Microsoft Corporation | Scalable music recommendation by search |
US20090277322A1 (en) * | 2008-05-07 | 2009-11-12 | Microsoft Corporation | Scalable Music Recommendation by Search |
US20090281906A1 (en) * | 2008-05-07 | 2009-11-12 | Microsoft Corporation | Music Recommendation using Emotional Allocation Modeling |
US8650094B2 (en) | 2008-05-07 | 2014-02-11 | Microsoft Corporation | Music recommendation using emotional allocation modeling |
US20100057734A1 (en) * | 2008-09-03 | 2010-03-04 | Yasushi Miyajima | Music processing method, music processing apparatus and program |
US8548960B2 (en) * | 2008-09-03 | 2013-10-01 | Sony Corporation | Music processing method and apparatus to use music data or metadata of music data regardless of an offset discrepancy |
US20100300271A1 (en) * | 2009-05-27 | 2010-12-02 | Microsoft Corporation | Detecting Beat Information Using a Diverse Set of Correlations |
US8878041B2 (en) * | 2009-05-27 | 2014-11-04 | Microsoft Corporation | Detecting beat information using a diverse set of correlations |
US8180765B2 (en) | 2009-06-15 | 2012-05-15 | Telefonaktiebolaget L M Ericsson (Publ) | Device and method for selecting at least one media for recommendation to a user |
US20100318544A1 (en) * | 2009-06-15 | 2010-12-16 | Telefonaktiebolaget Lm Ericsson (Publ) | Device and method for selecting at least one media for recommendation to a user |
US20120310925A1 (en) * | 2011-06-06 | 2012-12-06 | Dmitry Kozko | System and method for determining art preferences of people |
US8577876B2 (en) * | 2011-06-06 | 2013-11-05 | Met Element, Inc. | System and method for determining art preferences of people |
US20140143222A1 (en) * | 2012-11-16 | 2014-05-22 | Google Inc. | Ranking signals for sparse corpora |
US9779140B2 (en) * | 2012-11-16 | 2017-10-03 | Google Inc. | Ranking signals for sparse corpora |
US20190325894A1 (en) * | 2014-06-30 | 2019-10-24 | Dolby Laboratories Licensing Corporation | Perception based multimedia processing |
US10748555B2 (en) * | 2014-06-30 | 2020-08-18 | Dolby Laboratories Licensing Corporation | Perception based multimedia processing |
TWI584139B (en) * | 2016-09-09 | 2017-05-21 | Speech recognition system and its information processing method applied to non - real - time signal source | |
WO2023232865A1 (en) | 2022-06-03 | 2023-12-07 | Xandrie SA | Electronic system for digital and physical media |
Also Published As
Publication number | Publication date |
---|---|
US20040107821A1 (en) | 2004-06-10 |
US7081579B2 (en) | 2006-07-25 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US7081579B2 (en) | Method and system for music recommendation | |
US10152517B2 (en) | System and method for identifying similar media objects | |
US7797272B2 (en) | System and method for dynamic playlist of media | |
US6545209B1 (en) | Music content characteristic identification and matching | |
US8112418B2 (en) | Generating audio annotations for search and retrieval | |
US7240207B2 (en) | Fingerprinting media entities employing fingerprint algorithms and bit-to-bit comparisons | |
US20060206478A1 (en) | Playlist generating methods | |
US8005724B2 (en) | Relationship discovery engine | |
US8073854B2 (en) | Determining the similarity of music using cultural and acoustic information | |
US7745718B2 (en) | Music-piece classifying apparatus and method, and related computer program | |
US8519249B2 (en) | Music intelligence universe server | |
US7860862B2 (en) | Recommendation diversity | |
US7849092B2 (en) | System and method for identifying similar media objects | |
US20050038819A1 (en) | Music Recommendation system and method | |
US20030089218A1 (en) | System and method for prediction of musical preferences | |
US20020147628A1 (en) | Method and apparatus for generating recommendations for consumer preference items | |
US20090055376A1 (en) | System and method for identifying similar media objects | |
US20090254554A1 (en) | Music searching system and method | |
US20100063975A1 (en) | Scalable system and method for predicting hit music preferences for an individual | |
JP2005018205A (en) | Musical piece retrieving system and method, and musical piece retrieving device | |
US7254618B1 (en) | System and methods for automatic DSP processing | |
WO2002029610A2 (en) | Method and system to classify music |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
STCB | Information on status: application discontinuation |
Free format text: ABANDONED -- FAILURE TO RESPOND TO AN OFFICE ACTION |