1
USER-ORIENTED MULTIMEDIA
PRESENTATION SYSTEM FOR MULTIPLE
PRESENTATION ITEMS THAT EACH
BEHAVE AS AN AGENT
BACKGROUND OF THE INVENTION
The invention relates to a user-adaptive audio and/or video presentation system containing a background presentation sub-system for presenting audio and/or video items, control means for controlling said presentation sub-system as regarding an automatically generated selection sequence amongst said items according to a user preference pattern, and output means for physically presenting selected items to a user. Present-day delivery systems are quickly growing in terms of storage capacity and presentation bandwidth. Vis a vis a more or less constant absorption capacity for information flow in humans, this confronts the user person with an increasingly complex decision problem for choosing between various alternative presentations. A non-limiting example of such system is a jukebox loaded with Compact Discs, that could at present contain a hundred discs with some twenty songs each. The user could then choose along various strategies, such as random amongst the discs and sequential per disc, but this has been found too restrictive. On the other hand, random selection among all tracks present is felt as insufficiently coherent.
The selection problem sketched above can occur in various other environments, such as when choosing between a hundred or more concurrent TV channels, that each feature a sequence of items, or in a large collection of CD Video discs. Similar situations can occur with video clips or video games. The items may be based on uniform technology such as in the case of CD records. In a multimedia situation, the items may be intermixed, such as audio records competing against TV channels, or in the case of simultaneously selecting among still pictures as well as among audio items for concurrent presentation of the chosen audio plus video. Competing items need not have uniform presentations, such as an audio record versus an interactive audio-plus-videoplus-graphics game. The problem may occur on several hierarchical levels at a time, such as when simultaneously selecting among audio tracks as well as among whole audio records.
The present inventors have encountered a need for machine implementing a sequential choice amongst the items that is both meaningful in view of the user's preferences, but on the other hand is sufficiently nonuniform between different presentation sessions, and thereby gives both some coherence between presentations that are relatively close in time, but also certain startling effects through differentiating between such presentations and introducing a certain variability between different sessions.
SUMMARY TO THE INVENTION
Therefore, amongst other things it is an object of the invention to provide a system as recited above that gives more coherent presentation sequences than random selection alone would produce, but on the other hand observes more variability than what has been marketed up to now as 'favourite track selection': therein, after programming of a particular sequence, this sequence remains fixed. A particular additional problem of 'favourite track selection' is the need for individual programming by each user, which many users have felt as cumbersome. On the other hand, in 'shuffle play' the player itself executes a random selection amongst items that are physically available; in the view of the present
2
invention, a system of this kind has too little coherence in a sequence of items presented.
According to a first aspect of the invention it is thereto characterized in that each said item has a uniformly struc
5 tured first set of static attribute data and associated second set of weight value data, and also a third set of dynamic behaviour parameter data, and said items are with respect to their dynamic behaviour collectively governed by a set of behavioral and interactivity prescriptions, so that each
1° respective item represents a respective autonomous agent, and said control means include processing means for under influence of said first and second sets of data and as governed by said prescriptions, updating said weights and dynamic behaviour parameter data of each item in question
15 vis a vis other said items for subsequently influencing said controlling. The formulating as autonomous agents allows for a wide variability of presentations, while also retaining a certain coherence between successive items. The three levels for defining the properties of the items (attribute,
20 weight and actual behaviour) allow wide applicability of the principles explained hereabove.
By itself, earlier developments by the present assignee have led to other agent-based systems, such as disclosed in U.S. Pat. No. 5,418,887 (PHB33549), EP 704 077, corre
25 sponding U.S. patent application Ser. No. 08/418,995 (PHB 33903), EP Application 724 751 corresponding U.S. patent application Ser. No. 08/498,289 (PHB33914), EP Application 722 592, corresponding U.S. patent application Ser. No. 08/498,280 (PHB33915), and EP 95201526.1, correspond
30 ing U.S. patent application Ser. No. 08/655,169 (PHN15336).
Now, the present invention implements the items as autonomous agents. Such autonomous agents contain their
35 own static attribute data and dynamic weight data. A weight indicates the relevance of the associated static attribute. The dynamic behaviour parameters represent the actual behaviour of the item. The behavioral and interactivity prescriptions govern the evolving behaviour, under influence of
4Q static and dynamic properties of the item, and under influence of interactions between the respective items.
Advantageously, said behaviour is metaphored as motion within a finite space of at least two dimensions. This allows continual evolution of the behaviour of the items, and
45 thereby lends maximum flexibility thereto. Moreover, it allows an excellent visualization of the behaviour, thereby enabling all kinds of potential fine-tuning by a user, through effecting various selections. The behaviour can be visualized on a display screen, either at prototype or high-end machines
5q only, or alternatively, on all machines that have a display screen.
Advantageously, such system is arranged for forming among said items one or more temporary association clusters, which forming is conditional to similar behaviour
55 amongst the respective item agents in such cluster, and said presentation favours an item pertaining to a most recently accessed said cluster. In particular, the notion of a cluster has been found to introduce some structure among the various items, thereby giving a certain amount of coherence in the
60 presentation sequence.
Advantageously, the system has user evaluation input means for selectively influencing said updating at least with respect to said weight value data for a particular item through an evaluatory physical input. In this manner, opin
65 ions of an actual user can influence the behaviour of the agents and thus codetermine the outcome of the selection process.
3
Advantageously, a system having freeze means for blocking autonomy of said agents, and thus freezing said behaviour as dictated by said autonomy. For example, the freezing can become effective at the end of a presentation sequence. The same machine may then cater to various user preference patterns, that could differ between persons, as well as between different situations, such as depending on time-ofday. The freezed situation could be stored for later reactivation. The invention may be used in a network environment, either with broadcast presentation, or with personalized presentation, such as with earphones. Each user could have his own evaluation mechanism for influencing the agent behaviour.
Advantageously, said presentation sub-system is a storage system for storing and accessing audio items which are distributed amongst a plurality of physical media instances that each allow collective storage of a plurality of such items. This is a particularly advantageous embodiment. The invention also relates to a method for audio and/or video presentations along the principles set out above. Various further advantageous aspects are recited in dependent claims.
BRIEF DESCRIPTION OF THE DRAWING.
These and other aspects and advantages of the invention will be explained more in detail with reference to the detailed disclosure of preferred embodiments hereinafter, and in particular with reference to the appended Figures that show:
FIG. 1, a block diagram of an exemplary embodiment;
FIG. 2, an exemplary configuration of stored items;
FIG. 3, a sample display of a graphical user interface;
FIG. 4, a typical listening scenario of a system;
FIG. 5, a detailed object diagram of a set of items;
FIG. 6, a detailed object diagram of item behaviour;
FIG. 7, a detailed object diagram of item interaction;
FIG. 8, a detailed object diagram of bookkeeping;
FIG. 9, graphical symbols used in an object diagram;
FIG. 10, an illustration of the parameters Coverage and
Precision.
A FEW USER ASPECTS OF THE PRESENT
INVENTION
The invention has for the present embodiment been abbreviated as PATS (Personalized Automatic Track Selection). In the terminology of Compact Disc, usually each track corresponds to exactly one song or other audio equivalency. An earlier system for giving users influence on the presentation sequence in advance of the actual presentation, has been marketed as 'favourite track selection' (FTS). Therein, the user programs the disc player to follow a particular sequence amongst the tracks with respect to a particular Compact Disc or other medium. For players that allow a more or less random selection amongst a plurality of discs, such favourite tracks selection can be implemented both within a particular disc, and also amongst those various discs, in the way of a jukebox. Users have felt the programming a tedious effort. Further, the programming is generally linked to the memory of a particular player. Even if the programming sequence would be stored on a writable part of the disc, it is fixed. However, users may want to apply the present invention in a number of different activities, such as active listening, background music, and personalizing the sequence for use at a particular occasion, such as giving it as a present or playing it at a party.
4
The present invention serves user intentions that are not optimally addressed by current systems. Examples of such specific intentions that the present system tries to accommodate are:
5 people may enjoy somebody else doing the selecting of music for them, such as disc jockeys or radio programmes.
people like to listen to music that is known to them. 1Q people like a certain coherence between successive music tracks that form a listening session, such as thematic radio programmes or radio channels focusing on a particular style, people appreciate assistance in the selection process when j5 the amount of available music to choose from becomes larger, such as in the case of magazines or sampler CD's.
somebody who 'knows' the user can influence the selection process by presenting interesting or unexpected 20 options, such as would be the case for a sales person or an acquaintance.
the selection of certain music tracks is influenced by the ones previously listened to.
25 DETAILED DISCLOSURE OF PREFERRED
EMBODIMENTS
FIG. 1 is a block diagram of an exemplary embodiment of the invention, in particular as realized in the context of a
30 Compact Disc player with jukebox functionality. For reasons of clarity, all parts and subsystems have been shown in fragmentary manner only. Now, a stack 20 of records is rotated by means of a drive mechanism symbolized by axis 18. Access mechanism 22 allows to select among the various
35 records, and for each record among the tracks thereon. If applicable, both read and write on disc can be allowed, although the invention may in principle function with readonly. Player body 24 comprises various subsystems. First, block 28 controls the accessing by access mechanism 22. To
40 this effect, external controls 26 such as user buttons or remote control have been provided. Further control and selection signals are produced by processing means 34. For simplicity, blocks 28, 30, 34 have been linked in a chain; it is understood that blocks 34 and 28 may interact immedi
45 ately. The mechanical access facility is only one of many feasible embodiments as well.
Block 30 represents the audio processing; the audio output proper is symbolized by arrow 32, and the output is time-conformingly, so that the user experiences the resulting
50 speech, music, etcetera as a natural phenomenon. Of course, fast-forward and other modes can be provided as well, both for standard output, and with respect to adaptation to the user's wishes. Moreover, in the case of a (video) game, the presentation need not follow a uniform temporal sequence;
55 also still pictures could be presented. Parallel output of more than one presentation stream would be feasible as well, such as speech combined with background audio from an unconnected track from stack 20. Video and audio can be combined; animation from a game can be combined with audio
go of a kind preferred by the user.
Processing means 34 implement the dynamic behaviour of the items as agents. It may be based on conventional hardware, whereas the associated software will be discussed hereinafter. The software may be resident; the data that
65 relate directly to the items may be received from the audio record, but also from an external facility such as Internet. If received from the discs, this is done by an initial scan from
|
