WO2017179778A1

WO2017179778A1 - Search method and apparatus using big data

Info

Publication number: WO2017179778A1
Application number: PCT/KR2016/010324
Authority: WO
Inventors: 김인중
Original assignee: (주)솔루젠
Priority date: 2016-04-15
Filing date: 2016-09-13
Publication date: 2017-10-19
Also published as: KR101850853B1; KR20170118399A

Abstract

Provided are a method and an apparatus for searching for a keyword through an electronic terminal by using big data. A big data database to be searched is set in the electronic terminal. A search keyword is received through an input unit of the electronic terminal. The electronic terminal selects a document including the search keyword from among documents included in the big data database. The electronic terminal counts the number of words or phrases included in the selected document. The electronic terminal ranks the counted words or phrases in order of appearance frequency. The electronic terminal selects a plurality of ranked words or phrases as a plurality of related search words. The selected plurality of related search words are visually displayed on a display of the electronic terminal.

Description

Search method and device using big data

The present invention relates to a search method and apparatus using big data.

Recently, information processing technology has been rapidly developed to receive a query from a user in an information processing device such as a smart phone, a tablet PC, a personal digital assistant (PDA), and search for data corresponding to the user query in a plurality of databases. Research on search methods and systems provided to users is being actively conducted.

In particular, with the advent of the big data era, a content search engine system can quickly search for and provide accurate information corresponding to a user query in a plurality of databases in which a large amount of data is recorded. It is expected to be used in various fields such as intelligent robots, next generation PCs, telematics and home networks.

Although the concept of big data has been around for some time, the use of such big data has largely been an area where companies deal with individuals, companies deal with companies, or are used in the public sector.

On the other hand, the existing Internet keyword search, for example, by searching the data stored in the relevant server of the portal site or data on the Internet to find and present (or highly relevant) data containing the keyword. The materials presented are usually web pages, picture files, and video files. Except for pictures and videos, in most cases, links to homepages with text data are often found.

Conventional search finds specific material related to a search keyword (i.e., a particular press article, a particular blog post, a specific image, a specific video, etc.). Previously, the top search item (homepage link) in the search may or may not be representative of 'Hong Gil-dong'. In other words, it is only the most recent post, so it may appear at the top, and even if it is not a topic, it has been exposed to the top. This is a phenomenon that occurs because the existing search judges only that the search keyword is included and does not consider the contents of the search result at all.

According to the present invention, a method for searching a keyword using big data through an electronic terminal, comprising: setting a big data database to be searched in the electronic terminal; Receiving a search keyword through an input unit of the electronic terminal; Selecting, by the electronic terminal, a material including the search keyword among respective materials included in the big data database; Counting, by the electronic terminal, the number of words or phrases included in the selected data; Ranking, by the electronic terminal, the counted word or phrase in order of appearance frequency; Selecting, by the electronic terminal, the plurality of ranked words or phrases as a plurality of related search words; And visually displaying the selected plurality of related search words on a display of the electronic terminal.

Preferably, the displaying may be performed in larger letters as a related search word having a higher ranking.

Preferably, the displaying may be performed in a larger circle or a larger polygon box with a higher ranking search term.

Preferably, the displaying may be performed by changing the color of the text according to the ranking, or writing the letters of some ranks horizontally and writing the letters of the other ranks vertically.

Preferably, the big data database is an SNS article.

Preferably, a specific company or business is evaluated by the keyword search.

According to the present invention, there is provided an electronic device for searching a keyword using big data, comprising: a setting unit for setting a big data database to be searched for; An input unit to receive a search keyword; Control unit; And a display, wherein the controller is configured to select a material including the search keyword among respective materials included in the big data database, count the number of words or phrases included in the selected data, and The counted words or phrases are ranked in order of appearance frequency, the plurality of ranked words or phrases are selected as a plurality of related search terms, and the display is controlled to visually display the selected plurality of related search terms. A keyword retrieval electronic device is provided.

Preferably, the display control of the control unit displays the related search word having a higher ranking in larger letters.

Preferably, the display control of the control unit displays the higher-ranked related search word in a larger circle or a larger polygon box.

Preferably, the display control of the control unit is arranged to change the color of the text according to the ranking, or to write the letters of some ranks and to write the letters of some other ranks vertically.

Preferably, the big data database is an SNS article.

Preferably, the keyword search evaluates a particular company or business.

According to the search of the present invention, since the related search word is a vocabulary that is used most frequently with the search keyword, it provides a very accurate result for the search keyword.

Furthermore, in addition to merely listing the rankings in text, according to one embodiment of the present invention, the first related search word is displayed in the largest font, the second related search word is displayed in the next large font, and the third association is used. The search term can then be differentiated by displaying the next larger letter (ie, a smaller letter than the second associated search term). Other visual arrangements are also acceptable.

Through such big data search, looking at related search terms, it is easy to identify what references are coming and going on the Internet (or more narrowly on SNS) about the search keyword, and what is the issue about the search keyword. It does not require additional human effort as in the conventional search, and solves the conventional problem (that is, the problem of obtaining wrong information) in which garbage information (garbage or noise) can be searched at the top.

1 is a diagram visually showing an example of a search result according to the present invention.

2 shows an example of a flow diagram of a search according to the present invention.

Figure 3 shows the flow of the search according to the present invention divided into four stages.

4 to 6 are examples of making a search according to the present invention in the form of a smartphone app, and show a process of inputting a user request association word mentioned in FIG. 3.

7 to 8 are examples of making a search according to the present invention in the form of a smartphone app, and show a process of 'real time data collection' mentioned in FIG. 3.

9 to 10 are examples of making a search according to the present invention in the form of a smartphone app, and show the 'data analysis' process mentioned in FIG. 3.

FIG. 11 is an example of making a search according to the present invention in the form of a smartphone app, and illustrates the process of 'visualization' referred to in FIG. 3.

12 is a diagram illustrating collection of portal issue word association words.

13 shows an example of an apparatus according to the invention.

Preferably, the big data database is an SNS article.

Preferably, a specific company or business is evaluated by the keyword search.

Preferably, the big data database is an SNS article.

Preferably, the keyword search evaluates a particular company or business.

Hereinafter, a system and method according to the present invention will be described with reference to the drawings.

Existing Internet keyword searches, for example, search for data stored in relevant servers of portal sites or data on the Internet to find and present (or highly relevant) material containing the keyword. The materials presented are usually web pages, picture files, and video files. Except for pictures and videos, in most cases, links to homepages with text data are often found.

In the present invention, big data is used as a database, and the search results are not simply displayed in letters, but the related keywords are displayed by differentially increasing the size of the letters so as to be intuitively recognized.

For example, if a particular candidate's name is entered in an existing search site (Naver, Daum, Google, etc.) in relation to the election of the members of the parliament, homepage link materials such as newspaper articles about the person will be listed. However, according to the present invention, when a specific candidate name is input, the search database differs from the conventional one as big data. In addition, the web page is not searched, but is shown at the top based on the number of times mentioned in big data such as SNS. Show at the top is also possible to show at the top of the search results, but more preferably to show the largest of the most mentioned.

If a specific candidate name (Hong Gil-dong) is searched according to the present invention, various keywords may appear, but for example, Hong Gil-dong has been exposed to the press for military service corruption, and so on SNS, he has been mentioned much about the military corruption of candidates. Assume that, for the search keyword "Hong Gil-dong", the method and apparatus of the present invention may refer to the first related keyword (i.e., "military corruption" most frequently mentioned for candidate Hong Gil-dong on SNS as the related keyword (ie, the search result). The keyword 'Hong Gil-dong' and the keyword most frequently mentioned together in the SNS article) can be selected. For example, if an SNS article (Twitter Mention, etc.) says "Hong Gil-dong candidate is suspected of military service corruption", the related search term of "Hong Gil-dong" keyword is limited to this SNS article, "candidate" "military corruption" "suspect" The more articles with the purpose of this article (eg, "Hong Gil-dong is under investigation for military service corruption"), the more relevant the number (matches) of "military corruption" to the search keyword "Hong Gil-dong" It will get higher. And, if there are many SNS articles using similar vocabulary or the SNS is expanded / reproduced (eg, Twitter retweet function, etc.), the degree of association (matching number) will gradually increase.

Then, if a keyword frequently referred to together with the keyword 'Hong Gil-dong' is 'commit', 'commit' may be selected as the second related keyword. In this way, along with the keyword 'Hong Gil-dong', the keywords most frequently used in big data such as SNS articles are ranked. In this case, the search target is not limited to a post stored in a specific server such as a portal site such as Naver, but big data, and thus the accuracy of the search can be expected to be quite high. The criterion for selecting the related keyword may be, for example, limited to only the keyword mentioned together with the search keyword more than a thousand times and ranked among them.

From the above description, it can be seen that this is very different from the search in the existing portal site. Searches on existing portal sites may only target data pre-collected by the portal site (e.g. Naver, Daum), or may search the entire Internet, even if not previously collected (e.g., Google). Although the present invention can also target the general Internet as a search target, it is preferable to target only the big data, among others. Big data refers to SNS articles such as Twitter and Facebook as an example. Of course, in some cases, as described above, a search may be performed including the entire Internet data (for example, Internet press articles, blog posts, etc.). However, the difference from the existing search is that it does not search for specific material related to the search keyword (ie, specific press article, specific blog post, specific image, specific video, etc.), but searches the entire big data. It finds the frequency of other related keywords that appear with the keyword (that is, in the same paragraph or in the same article), ranks them, and displays the ranking. Previously, the top search item (homepage link) in the search may or may not be representative of 'Hong Gil-dong'. In other words, it is only the most recent post, so it may appear at the top, and even if it is not a topic, it has been exposed to the top. This phenomenon occurs because the existing search judges only that the search keyword is included and does not consider the contents of the search result at all. However, according to the search of the present invention, since the first related search word is a vocabulary that is used most frequently with the search keyword, it provides a very accurate result for the search keyword.

Furthermore, in addition to merely listing the rankings in text, according to one embodiment of the present invention, the first related search word is displayed in the largest font, the second related search word is displayed in the next large font, and the third association is used. The search term can then be differentiated by displaying the next larger letter (ie, a smaller letter than the second associated search term). In this case, the first related search word does not necessarily need to be at the top, but may appear in the center of the search terminal screen, or may be displayed in other positions. Since the font size is the largest, it can be easily understood that the first related search word is displayed anywhere on the screen.

As such, the font size may also indicate the degree of relevance ranking, or the degree of relevance ranking may be indicated by the size of the circle so as to put the first related search term in the largest circle and the second related search term in the next large circle. Of course, you can use the criteria of the size of the circle and the size of the letter together. The position of the circle can also be displayed at the top of the largest and gradually smaller toward the bottom, the largest can be displayed at the leftmost and gradually smaller toward the right, or randomly displayed on the screen and the size can be viewed by the user. You can make a judgment. The text does not necessarily need to be written horizontally, and some related search words may be written horizontally and some related search words may be written vertically so that they can be visually and intuitively understood. In addition, the color may be changed for each related search word.

Through such big data search, looking at related search terms, it is easy to identify what references are coming and going on the Internet (or more narrowly on SNS) about the search keyword, and what is the issue about the search keyword. have.

On the other hand, not only the listing of the plurality of related search words, for example, when the first related search word is clicked on the screen of the search terminal, it is also possible to provide a service that further shows in which phrase the first related search word appeared.

That is, if the first related search term is "military corruption" and the second related search term is "commitment" using "Hong Gil-dong" as the search keyword, "commitment" is the largest letter on the search result screen, and then "commitment" is next. The large text and the rest of the related search terms will be listed in the appropriate font size for each ranking. Each related search term can be configured to be clickable, and when a specific related search term (eg, "commitment") is clicked, a specific article about which "Hong Gil-dong" and "commitment" are mentioned together can be viewed. It is. The specific writing is a writing that constitutes a part of the big data, and may be a normal SNS writing, but may be searched to include a writing of a normal internet site (eg, an internet press or a blog) as necessary.

In the above, the lawmaker was taken as an example, but another example is shown in FIG. The principle of operation is the same as described above.

That is, the most related word (first related search word) for the search term 'tuya' is “Kim Ji-hye” and is indicated in large letters in the center of FIG. 1. The next most relevant word (second related search term) is "Sugar Man", which is indicated in large letters at the bottom slightly in the middle of FIG. The next most related word (third related search term) is "eye glasses", which are indicated in a slightly smaller font at the top slightly in the middle of FIG. It can be seen that other related search terms (ie, below the fourth related search term) are displayed in a form enclosing the letter "Kim Ji-hye" in the center.

In the service of the method or apparatus according to the present invention, the general service version may provide only related search terms, and the paid service version may be configured so that a specific article (which phrase comes from) may be clicked on.

Big data may be provided through, for example, an association / agreement with a company such as Teradata. This linkage will enable individuals to see the big data of public services.

As another example, for example, a search for a specific restaurant name may be assumed. According to the present invention, when searching for a specific restaurant name, for example, a related search term "weave" may come out, and a related search term "delicious" may come out. Both of them appear a lot, but the most relevant search term is "weave". That is, when a specific restaurant name is searched, if the first related search term is "weave" and the second related search term is "delicious," the user of the present invention may cook the salty food while the specific restaurant is delicious, but is generally cooked. Can be obtained. According to the existing search, there is no choice but to read the reviews of the restaurant that stands at the top. This is a time consuming task and the review is not necessarily fair. In other words, in a conventional search where a promotional blog post may be searched at the top of a coincidence, it is very likely to have wrong information about a particular restaurant. However, according to the search of the present invention, for example, since there are a lot of articles (materials) in which the name of the restaurant and the word "weave" appear at the same time, it is possible to obtain a completely different effect from the conventional search in which one article is read in pieces. It can be obvious.

In particular, in the case of using the graphic in the present invention, for example, by varying the size of the circle to indicate the size of the degree of association, or to provide a variety of visual and intuitive interface through horizontal writing, vertical writing, color mixing, etc. You can get easy and accurate information.

According to this invention, as a result, it is possible for an individual to evaluate a restaurant, a company, or the like.

In the conventional search, the contents must be checked by reading the article with the human eye through the keyword search, and eventually, a large amount of human power is required. However, according to the present invention, since the machine filters the most frequently read letters by reading, the human power is not necessary. And, if a large number of searches are displayed in large letters, it is very easy to grasp the degree of association, and discrimination is given. Not only large letters but also highlights, color differences, or any other visual effects may be given or combined.

Big data according to the present invention can be variously diversified, such as articles, blogs, SNS, Twitter. Some or all of these can be used and can be applied to new, separate database systems as needed.

The analysis for searching (selecting related search terms) of the present invention may use techniques such as machine learing and deep learning as a big data analysis technique. The method and apparatus of the present invention are based on the web, Or it may be implemented based on the app (application) of the smartphone.

On the other hand, it is possible to click again on a part of the search result (i.e., display related search terms) to search again, or to link to Twitter or blog. Re-search means that when "Hong Gil-dong" is searched and "military corruption" is the first related search word, when "military corruption" is clicked, "military corruption" now becomes a search keyword and searches for the related search word accordingly.

The term 'associated search term' itself may be a somewhat broad term, but it should not be understood simply by the meaning of the word, but the preconditions of the present invention as described above, that is, the form of the show, the form of the app, the search Since the term refers to a related search word within a condition that conditions, such as object or method, are satisfied, the term is not judged solely.

That is, according to the prior art, it is impossible or difficult for a service that an individual can evaluate against a company or a mart, but according to the present invention, each of the words constituting the big data becomes a meaningful evaluation material, and a specific word is simply It's easy to see that it's a different way of searching and showing what you've entered. For example, in the related art, if a new credit card is issued, the service is merely searched and evaluated internally. However, according to the present invention, it is obvious that the service can be used in a completely different manner as a service that allows the general public to view big data.

According to the present invention, as one application, if a search term "Hong Gil-dong" is put in, a quiz play for predicting what the first related search term may appear may be played. Or you can play a quiz that predicts when a search term will appear "military irregularities" when you enter. It may also function as a simple and fun play among young people.

In the conventional search, although the amount of data is large, in fact, there is a difficulty in access and processing. Even if you do it in a PC, not a smartphone, it is the same. For example, in order to avoid such troublesomeness (that is, searching for useless information), there is a person who enters and searches the homepage of a newspaper, and for the person who wants to filter out the noise, the present invention can be a good tool.

Currently, newspaper articles on the Web often place search results in order of importance. However, as you go through the first article to the last, it seems that more and more similar words are used, so it's hard to know the subtleties before a person sees each article. However, according to the present invention, since the words of each article (when the big data is a newspaper article database) are counted, it has a new meaning different from the conventional search. That is, the central keyword may be selected according to the frequency of occurrence of the word. The central keyword may be regarded as the same concept as the aforementioned related search word.

Meanwhile, the specific form of the visualization has been described above, but the present invention is not limited thereto, and may provide discrimination in various visual forms.

The word count (ie association checking) process may use various algorithms and weighting methods. If some data in the big data is determined to be a commercial article, the article may be excluded from the word count. Determination of the commerciality may be possible by various means, for example, by checking whether the text of the material contains a predetermined (predetermined) propaganda text, or by checking whether a specific homepage address is included. .

The present invention provides a method and apparatus for searching for keywords using big data through an electronic terminal.

For example, an electronic terminal sets a big data database to be searched for (step 201). Then, the search keyword is input through the input unit of the electronic terminal (step 202).

The electronic terminal selects the data including the search keyword from among the materials included in the big data database (step 203). The electronic terminal counts the number of words or phrases included in the selected data (step 204). Then, the electronic terminal ranks the counted words or phrases in order of appearance frequency (step 205).

The electronic terminal selects a plurality of ranked words or phrases as a plurality of related search terms (step 206). The selected plurality of related search words are visually displayed on the display of the electronic terminal (step 207).

The present invention may operate in a PC connected to a big data database and a network, or may operate in a device such as a smartphone.

The present invention can be viewed as being largely passed through a 'user request association word input' process, a 'real time data collection' process, a 'data analysis' process, and a 'visualization' process.

Each of these will be described below.

In the upper part of FIG. 4, a search box (input box) is displayed, and in the lower part, 20 real-time search words that are an issue in a portal site (eg, Naver, Daum, etc.) are displayed by rank.

For example, the main page of the app according to the present invention is written in HTML format for the purpose of a web app. In an HTML document, you configure a user's action as a java script to pass values to an internal program.

Through the input window displayed in the upper part of FIG. 4, the user can receive a word that he / she wants to find, and the user can click the enter key or the analysis button below to transfer the input content to the internal collection program.

Referring to the lower part of FIG. 4, the top 20 list windows of real-time search terms of portal sites are arranged at the lower part separately from the user input part of the upper part of the page. It automatically collects real-time search term ranking, collects and analyzes web data related to the word, and places it on the main screen where the user accesses to add additional value for user convenience.

The user request association word function (ie, a function related to the input window in the upper part of FIG. 3) provides a function of allowing a user to enter at least one to many desired words by accessing the main page of the app according to the present invention. . Technically, text can be received as an input value, and multiple words can be passed as input values separated by spaces. If you press the Analyze button or click Enter without entering a word, you are prompted to enter a word. This may be implemented through, for example, the code shown in FIG. 5.

Next, when the user enters a normal word and clicks the Analyze button or Enter key, the internal search.jsp file is called to deliver the received argument value. search.jsp passes the passed argument values back to the data collection program. This may be implemented, for example, through the code shown in FIG. 6.

Collecting real-time data on the Internet runs a web scraping engine, searches for documents containing the words entered by the user, and stores the results in html or plain text. The search result uses the word entered by the user as the file name. This may be implemented through, for example, the code shown in FIG. 7.

The initial data collected through the web scraping engine contains html tags and unnecessary characters that cannot be analyzed immediately, so data filtering is performed to clean up these unnecessary contents. Once the filtering is done, the file will be in the form of a plain text file. This may be implemented through, for example, the code shown in FIG. 8.

In detail, as shown in FIG. 4, the present invention collects data from various sources such as web, SNS, blog, and the like to find a related word of a user request (that is, a word entered in the input window of FIG. 4), or a portal. An association word of an issue word (ie, words marked as issues 1 to 10 and issues 11 to 20 at the bottom of FIG. 4) may be collected.

Data analysis enables Korean parsing using KoNLP package of open software R. It takes a text file that has been filtered and deletes unnecessary symbols that could not be removed by the filtering program. This may be implemented through, for example, the code shown in FIG. 9.

Next, we rewrite the extracted nouns into a list and measure the frequency of each noun. Output the top 100 words in a separate csv file for secondary analysis of nouns and noun frequencies. Extract 20 of the top 100 words you output and use them as keywords in your search results. This may be implemented through, for example, the code shown in FIG. 10.

The visualization uses R's word cloud package to output in png file format. In the visualization of big data analysis results, the word cloud places the word with the highest frequency in the center and outputs the word with the smaller frequency. The more output the parameters are, the closer they are to the prototype. This may be implemented through, for example, the code shown in FIG. 11.

The result of completing the visualization is transferred to the first requested user screen. The result delivered includes the resulting png file and a list of the top 10 words. The search.jsp file creates the received png file name and word list in html format and delivers it to the user's screen. The top ten words of the analysis result are linked to the search box of the portal site (eg, Naver) for the user's re-search convenience.

By going through this process, for example, inputting "tuya" into the input window of Figure 4, it is possible to obtain a search result as shown in FIG.

Meanwhile, referring to the words of the issues 1 to 10 and the issues 11 to 20 shown at the bottom of FIG. 4, the list of the predetermined number of issue words (eg, 20) displayed on the main screen is an example every 5 minutes. Every time the portal automatically connects to the word list has been collected. In order to maintain the real-time information of the information provided, the same word is newly collected every two hours. The collected list is delivered to the web scrap program and the analysis result is stored in the server and provided to the user. This may be implemented through, for example, the code shown in FIG. 12.

13 shows an example of an apparatus according to the invention.

An electronic device for searching a keyword using big data is provided. The apparatus includes a setting unit 301 for setting a big data database to be searched, an input unit 303 for receiving a search keyword, a controller 305 and a display 307.

The controller 305 selects a material including a search keyword from among materials included in the big data database, counts the number of words or phrases included in the selected data, and counts the counted words or phrases. The ranking may be performed in order, a plurality of ranked words or phrases may be selected as a plurality of related search terms, and the display 307 may be controlled to visually display the selected plurality of related search terms.

The controller 305 may be, for example, a CPU of a PC or a CPU of a smartphone.

Although specific examples have been described above, the present invention is not limited to the above embodiments, and many modifications are possible within the essential spirit of the present invention as set forth in the appended claims by those skilled in the art. Of course. Various other modifications should be considered to be within the scope of the present invention, without departing from the basic spirit of the invention.

The present invention relates to a search method and apparatus using big data, and has industrial applicability.

Claims

A method for searching for keywords using big data through an electronic terminal,

Setting a big data database to be searched in the electronic terminal;

Receiving a search keyword through an input unit of the electronic terminal;

Selecting, by the electronic terminal, a material including the search keyword among respective materials included in the big data database;

Counting, by the electronic terminal, the number of words or phrases included in the selected data;

Ranking, by the electronic terminal, the counted word or phrase in order of appearance frequency;

Selecting, by the electronic terminal, the plurality of ranked words or phrases as a plurality of related search words; And

Visually displaying the selected plurality of related search terms on a display of the electronic terminal;

Keyword search method comprising a.
The method of claim 1,

The displaying step is a keyword search method for displaying the related words having higher ranking in larger letters.
The method of claim 1,

The displaying may include displaying the related search word having a higher ranking in a larger circle or a larger polygon box.
The method of claim 1,

The displaying may include a keyword retrieval method of changing the color of the text according to the ranking, or writing the letters of some ranks horizontally and writing the letters of some other ranks vertically,
The method of claim 1,

The big data database is an SNS article, keyword search method.
The method according to any one of claims 1 to 5,

A keyword search method for evaluating a specific company or business by the keyword search.
An electronic device for searching a keyword using big data,

A setting unit for setting a big data database to be searched for;

An input unit to receive a search keyword;

Control unit; And

display

Including;

The controller selects a material including the search keyword from among materials included in the big data database, counts the number of words or phrases included in the selected data, and counts the counted words or phrases. A keyword search electronic device that ranks the order of appearance frequency, selects the plurality of ranked words or phrases as a plurality of related search terms, and controls the display to visually display the selected plurality of related search terms. .
The method of claim 7, wherein

The display control of the control unit is such that the high-ranked related search word is displayed in large letters.
The method of claim 7, wherein

The display control of the controller is such that a higher ranking related search word is displayed in a larger circle or a larger polygon box.
The method of claim 7, wherein

The display control of the control unit may be configured to change the color of the text according to the rank, or to write the letters of some ranks horizontally and to write the letters of the other partial ranks vertically.
The method of claim 7, wherein

The big data database is an SNS article, keyword search electronic device.
The method according to any one of claims 7 to 11,

A keyword search electronic device for evaluating a specific company or business by the keyword search.