JP2008269427A - Information retrieval device, information retrieval method, information retrieval program, and computer readable recording medium recorded with information retrieval program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To display contents in an order reflecting a user's taste. <P>SOLUTION: The contents including a retrieval word for retrieving a content and a content identifier corresponding to the content are retrieved from a content storing part 301, a sentence adaptation degree indicating a frequency in which the content retrieved by the retrieval word appears is calculated, a content identifier having the same content identifier with the retrieved content identifier is retrieved from a browsing history log 302, a browsing popularity degree indicating the frequency of browsing content of the retrieved content identifier is calculated, and a content identifier retrieved on the basis of a result obtained by combining the calculated sentence adaptation degree and browsing popularity degree is outputted. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a technology for searching for content.

  Conventionally, in an information search system provided on the Internet, contents such as sentences and drawings are accumulated in a database, contents related to a search keyword given by a user are taken out from the database, and presented as search results to a user. Technology was used.

  When presenting the content that is the search target, usually, the text relevance between the search formula and each content of the search result is obtained based on the frequency of use of each search term included in the search formula, and the text related A method of displaying contents in descending order is used (see Non-Patent Document 1).

  However, even if the search term is frequently used, the content is not necessarily expected by the user. Conversely, even content that is used infrequently may be expected content. Therefore, the search results ordered using the text relevance level do not match the search results displayed in the order desired by the user, and it takes a lot of trial and error and time until the desired content is reached. There was a problem of giving to users.

Therefore, in Patent Document 1 and Patent Document 2, there is a technique for displaying the content requested by the user at a higher level by recording the number of browsing requests made by the user and correcting the above-described sentence relevance using this number of times. It is disclosed. As a result, content with a high browsing request is displayed at a higher level, and content with a low browsing request is displayed at a lower level.
JP 2005-190453 A JP 2006-338100 A Kitakenji, “Information Retrieval Algorithm”, Kyoritsu Shuppan, 2002

  However, a variety of languages are often used, and content that is frequently searched in close contact with daily life, such as weather forecasts and national place names, is a single search term entered by the user. However, if the content is included, the content is always prioritized and presented, so that the content expected by the user is not always displayed preferentially.

  In addition, even if a specific user intentionally browses a large amount, it is reflected in the number of browsing requests, so unlike the original browsing request, the intentional content is displayed with priority. there were.

  This invention is made in view of the above, and makes it a subject to display the content reflecting a user's taste.

  The present invention according to claim 1 is a content storage means for storing a content and a first content identifier that can identify the content in association with each other, and a browsing for storing a second content identifier of the browsed content. A history storage means, a search word receiving means for receiving a search word for searching for content, and the content and the first content identifier are read from the content storage means, and the content containing the search word and the content The first content identifier is searched, the content search means for calculating the sentence matching degree indicating the frequency at which the search term appears in the content, and the second content identifier is read from the browsing history storage means and searched. A second content identifier having the same content identifier as the first content identifier is detected. A first browsing history search unit that calculates a browsing popularity degree indicating a frequency of browsing the content of the first content identifier, a matching degree combining unit that combines the sentence matching degree and the browsing popularity degree, And a search result output means for outputting the searched first content identifier based on the result.

  In the present invention, the text matching indicating the frequency of appearance in the content searched for by searching the content storing means for the content including the search word for searching the content and the content identifier corresponding to the content. A content identifier having the same content identifier as the searched content identifier is searched from the browsing history storage means, and a browsing popularity degree indicating the frequency of browsing the content of the searched content identifier is calculated and calculated. Since the retrieved content identifier is output based on the result of combining the sentence suitability and the browsing popularity, the content reflecting the user's preference can be displayed.

  In the present invention according to claim 2, the browsing history storage means stores the second content identifier in units of rows, and the browsing popularity is the same content as the searched first content identifier. The gist is that the calculation is performed using the number of rows of the second content identifier having the identifier.

  The present invention according to claim 3, wherein the matching degree synthesizing unit calculates a score obtained by synthesizing the sentence matching degree and the browsing popularity degree, and the search result output unit has a high score. The gist is to output the first content identifier in order.

  In the present invention, the matching degree synthesizing unit calculates a score obtained by synthesizing the sentence matching degree and the browsing popularity degree, and the search result output unit outputs the content identifiers in the order of the highest score. Therefore, it is possible to preferentially provide a user with a popular content identifier that has a high degree of matching with a search term input by the user.

  The present invention according to claim 4, wherein the browsing history storage means stores the second content identifier in association with a browsing time when the content of the second content identifier is browsed. The first browsing history search means, after searching for the second content identifier, reads the browsing time corresponding to the second content identifier from the browsing history storage means, and indicates the elapsed time from the browsing time to the present. The gist of the invention is to further include a calculating means for calculating the browsing freshness and an adding means for adding the browsing freshness to the browsing popularity.

  In the present invention, since the browsing freshness indicating the elapsed time from the time when the same content identifier as the content identifier searched by the search word is browsed to the present is added to the browsing popularity, the original popularity degree is set. The reflected content identifier can be provided to the user.

  In the present invention described in claim 5, the browsing history storage means includes the second content identifier, the browsing time when the content of the second content identifier was browsed, and the user identifier of the user who browsed the content. The first browsing history search means searches for the second content identifier, and then stores the browsing time and the user identifier corresponding to the second content identifier in the browsing history. Read from the storage means, create a list A showing the relationship between the browsing time and the user identifier, and for each user identifier of the list A, the elapsed time from the browsing time to the present A calculation means for calculating the browsing freshness to be displayed, and the browsing freshness for all user identifiers in the list A are totaled, and the total browsing freshness is calculated as the browsing Adding means for adding the air permeability, it is summarized as further comprising a.

  In the present invention, since the browsing freshness indicating the elapsed time from the time when the same content identifier as the content identifier searched by the search word is browsed to the present is added to the browsing popularity, the original popularity degree is set. The reflected content identifier can be provided to the user.

  In the present invention according to claim 6, the calculating means reads the latest browsing time from the browsing history storage means among the browsing times corresponding to the second content identifier, and calculates the browsing freshness. Is the gist.

  In the present invention, since the freshness of browsing is calculated using the newest browsing time, it is possible to provide the user with a content identifier that more reliably reflects the original popularity.

  The gist of the present invention described in claim 7 is that the browsing freshness value is a smaller value as the elapsed time is longer, and a larger value as the elapsed time is shorter.

  The present invention according to claim 8 further includes user attribute storage means for associating and storing a user identifier of the user and the attribute of the user, wherein the browsing history search unit includes When an attribute is received, the user identifier and the attribute are read from the user attribute storage means, a user identifier having the same attribute as the received attribute is searched, and the searched user identifier and Creating means for creating a list B indicating a relationship with the attribute; and deleting means for deleting a user identifier not stored in the list B and a browsing time corresponding to the user identifier from the list A. In addition, the gist is to calculate the browsing freshness using the list A after deletion.

  In the present invention, a user identifier having the same attribute as the received user attribute is searched from the user attribute storage means, and a list B indicating the relationship between the searched user identifier and the attribute is created. In order to calculate the browsing freshness using the list A in which the user identifiers not stored in the list B and the browsing time corresponding to the user identifiers are deleted from the list A that has already been created, Content that is popular with nearby users can be preferentially presented.

  The present invention according to claim 9 is the first browsing history search unit, when the total number of user identifiers in the list A after deletion is less than a predetermined threshold, A gist is that a user identifier corresponding to a content identifier that matches a part is read from the browsing history storage unit, and the read user identifier and a browsing time corresponding to the user identifier are added to the list A. To do.

  In the present invention, when the total number of user identifiers in the list A is smaller than a predetermined threshold, the user identifier corresponding to the content identifier that matches a part of the content identifier searched by the content search means is viewed. Since reading from the history storage means and adding the read user identifier and the browsing time corresponding to this user identifier to the list A, the browsing freshness can be calculated using more related browsing information. .

  According to a tenth aspect of the present invention, the browsing history storage means further stores a browsing time when the content of the second content identifier is browsed, and the calculation means stores the list A in the list A. The gist is to read the browsing time corresponding to each user identifier from the browsing history storage means and calculate the browsing freshness by further using the browsing time.

  In the present invention, the browsing time corresponding to each user identifier in the list A is read from the browsing history storage means, and the browsing freshness is calculated by further using this browsing time. The reflected content identifier can be provided to the user.

  The present invention as set forth in claim 11, when the output first content identifier is selected by a user, content identifier receiving means for receiving the first content identifier, and content of the first content identifier From the acquired content acquisition means, the content analysis unit that analyzes the related content in the content directly related to the content from the acquired content, and extracts the related content identifier of the related content, and the browsing history storage means A second browsing history for reading the second content identifier, searching for a second content identifier having the same content identifier as the related content identifier, and calculating a browsing popularity indicating the frequency of browsing the content of the related content identifier Based on the search means and the browsing popularity, A content processing means for processing a serial content, and content output means for outputting processed the contents were, and summarized in that further comprising a.

  In the present invention, for the content with the content identifier selected by the user, the related content in the content directly related to the content is analyzed, and the browsing popularity for the analyzed related content is calculated. Since the content processed based on the browsing popularity is output, it is possible to easily search for related content that is popular in the output content.

  In the present invention according to claim 12, the second browsing history search means calculates the browsing popularity and extracts a related content identifier of related content outside the content indirectly related to the content. The content output means further outputs a related content identifier outside the content.

  In the present invention, since the related content identifier outside the content indirectly related to the content is further output, the convenience of the content can be improved.

  The present invention described in claim 13 is characterized in that the browsing history storage unit is a transition source content identifier of a transition source content that has been browsed before the content of the second content identifier stored in the browsing history storage unit is browsed. The second browsing history search unit reads the second content identifier from the browsing history storage unit, and stores a content identifier having the same content identifier as the received first content identifier. Search and create a transition source content identifier list in which the transition source content identifier corresponding to the content identifier is associated with the number of the transition source content identifiers, and the same transition source as the transition source content identifier in the transition source content identifier list Whether the transition source content identifier having the content identifier is the browsing history storage means A transition destination content identifier list is created by associating the searched transition source content identifiers with the number of the transition source content identifiers, and the transition is performed using the transition source content identifier list and the transition destination content identifier list. The gist is to calculate the importance of the destination content identifier and extract the transition destination content identifier as a related content identifier outside the content.

  The present invention according to claim 14 is the transition source content identifier of the transition source content browsed before the browsing history storage unit browses the content of the second content identifier stored in the browsing history storage unit. The second browsing history search unit reads the second content identifier from the browsing history storage unit, and stores a content identifier having the same content identifier as the received first content identifier. A list in which the transition source content identifier corresponding to the content identifier and the number of the transition source content identifiers are associated is created as a transition destination content identifier list, and the related content in the content is retrieved from the transition destination content identifier list. Delete the identifier and transition to the destination content identifier list after deletion Extracts a content identifier as the content out of the associated content identifier, and summarized in that the number of the transition destination content identifier and importance.

  The gist of the present invention described in claim 15 is that the content output means outputs a related content identifier outside the content according to the importance and / or together with the importance.

  In the present invention, the related content outside the content is output according to the importance and / or together with the importance, so that the popular content can be browsed more efficiently.

  According to a sixteenth aspect of the present invention, there is provided browsing history storage means for storing a second content identifier of browsed content, content identifier receiving means for receiving a first content identifier for desired content, and the first A content acquisition unit that acquires content with a content identifier, a content analysis unit that analyzes related content in content directly related to the content from the acquired content, and extracts a related content identifier of the related content; The second content identifier is read from the browsing history storage means, a second content identifier having the same content identifier as the related content identifier is searched, and a browsing popularity indicating the frequency of browsing the content of the related content identifier is calculated. Browsing history search means to Based on viewing popularity to a content processing means for processing the content, a content output means for outputting processed the contents were, summarized in that with.

  In the present invention, the content identifier for the desired content is received, the related content in the content directly related to the content is analyzed, the browsing popularity for the analyzed related content is calculated, and the browsing popularity Since the content processed based on is output, the related content that is popular in the output content can be easily searched.

  In the present invention described in claim 17, the browsing history search means calculates the browsing popularity and extracts a related content identifier of related content outside the content indirectly related to the content. The content output means further outputs related content identifiers outside the content.

  In the present invention, since the related content identifier outside the content indirectly related to the content is further output, the convenience of the content can be improved.

  The present invention according to claim 18 is the transition source content identifier of the transition source content that was browsed before the browsing history storage unit browsed the content of the second content identifier stored in the browsing history storage unit. The browsing history search unit reads the second content identifier from the browsing history storage unit and searches for a content identifier having the same content identifier as the received first content identifier. Creating a transition source content identifier list in which the transition source content identifier corresponding to the content identifier and the number of the transition source content identifiers are associated, and the same transition source content identifier as the transition source content identifier in the transition source content identifier list The transition source content identifier having And creating a transition destination content identifier list in which the retrieved transition source content identifiers and the number of the transition source content identifiers are associated, and using the transition source content identifier list and the transition destination content identifier list, The gist is to calculate the importance of the content identifier and extract the transition destination content identifier as a related content identifier outside the content.

  The present invention described in claim 19 is the transition source content identifier of the transition source content that was browsed before the browsing history storage unit browsed the content of the second content identifier stored in the browsing history storage unit. The browsing history search unit reads the second content identifier from the browsing history storage unit and searches for a content identifier having the same content identifier as the received first content identifier. A list in which the transition source content identifier corresponding to the content identifier and the number of the transition source content identifiers are associated is created as a transition destination content identifier list, and a related content identifier in the content is obtained from the transition destination content identifier list. Delete the transition destination code in the transition destination content identifier list after deletion. Extracts the Ceiling identifier as the content out of the associated content identifier, and summarized in that the number of the transition destination content identifier and importance.

  The gist of the present invention described in claim 20 is that the content output means outputs a related content identifier outside the content according to the importance and / or together with the importance.

  In the present invention, the related content outside the content is output according to the importance and / or together with the importance, so that the popular content can be browsed more efficiently.

  According to a twenty-first aspect of the present invention, a step of associating a content with a first content identifier that can identify the content is stored in the content storage means, and a second content identifier of the browsed content is stored in a browsing history. Storing in the means, receiving a search word for searching for content, reading the content and the first content identifier from the content storage means, and corresponding to the content including the search word and the content Searching for the first content identifier, calculating a sentence suitability indicating the frequency with which the search term appears in the content, and reading the second content identifier from the browsing history storage means, Second content identification having the same content identifier as the first content identifier Based on the result of combining, the step of calculating the browsing popularity indicating the frequency of browsing the content of the first content identifier, the step of combining the sentence suitability and the browsing popularity And a step of outputting the searched first content identifier.

  According to a twenty-second aspect of the present invention, the step of storing the second content identifier of the browsed content in the browsing history storage means, the step of receiving the first content identifier for the desired content, and the first content From the acquired content, analyzing the related content in the content directly related to the content from the acquired content, extracting the related content identifier of the related content, from the browsing history storage means Reading the second content identifier, searching for a second content identifier having the same content identifier as the related content identifier, and calculating a browsing popularity indicating the frequency of browsing the content of the related content identifier; Based on browsing popularity, A step of processing the Ntsu, and outputting the processed the contents were, and summarized in that with.

  According to a twenty-third aspect of the present invention, a process of associating a content with a first content identifier that can identify the content and storing the content in a content storage unit, and storing a second content identifier of the browsed content in a browsing history A process for storing in the means, a process for receiving a search word for searching for content, and reading out the content and the first content identifier from the content storage means to correspond to the content including the search word and the content The first content identifier to be searched, a process for calculating a sentence suitability indicating the frequency of occurrence of the search term in the content, and the second content identifier is read from the browsing history storage means, and the searched Search for a second content identifier having the same content identifier as the first content identifier, and The process of calculating the browsing popularity indicating the frequency of browsing the content of one content identifier, the process of combining the sentence matching degree and the browsing popularity, and the search result based on the combined result The gist is to cause a computer to execute a process of outputting one content identifier.

  The present invention as set forth in claim 24, the process of storing the second content identifier of the browsed content in the browsing history storage means, the process of receiving the first content identifier for the desired content, and the first content A process of acquiring the content of the identifier, a process of analyzing the related content in the content directly related to the content from the acquired content, extracting the related content identifier of the related content, and the browsing history storage means A process of reading the second content identifier, searching for a second content identifier having the same content identifier as the related content identifier, and calculating a browsing popularity degree indicating a frequency of browsing the content of the related content identifier; Processing to process the content based on browsing popularity , And summarized in that to execute a process of outputting the processed the contents were, to the computer.

  The present invention described in claim 25 relates to a process of associating a content with a first content identifier that can identify the content and storing it in the content storage means, and storing a second content identifier of the browsed content as a browsing history. A process for storing in the means, a process for receiving a search word for searching for content, and reading out the content and the first content identifier from the content storage means to correspond to the content including the search word and the content The first content identifier to be searched, a process for calculating a sentence suitability indicating the frequency of occurrence of the search term in the content, and the second content identifier is read from the browsing history storage means, and the searched Search for a second content identifier having the same content identifier as the first content identifier, and The process of calculating the browsing popularity indicating the frequency of browsing the content of one content identifier, the process of combining the sentence matching degree and the browsing popularity, and the search result based on the combined result The gist is to cause a computer to execute a process of outputting one content identifier.

  The present invention as set forth in claim 26, a process of storing the second content identifier of the browsed content in the browsing history storage means, a process of receiving the first content identifier for the desired content, and the first content A process of acquiring the content of the identifier, a process of analyzing the related content in the content directly related to the content from the acquired content, extracting the related content identifier of the related content, and the browsing history storage means A process of reading the second content identifier, searching for a second content identifier having the same content identifier as the related content identifier, and calculating a browsing popularity degree indicating a frequency of browsing the content of the related content identifier; Processing to process the content based on browsing popularity , And summarized in that to execute a process of outputting the processed the contents were, to the computer.

  According to the present invention, it is possible to display content reflecting the user's preferences.

[First Embodiment]
FIG. 1 is a configuration diagram showing a configuration of an information search system 1 according to the first embodiment. The information search system 1 according to the present embodiment includes a user terminal 2 and an information search device 3 and is connected to each other via a network 4.

  The user terminal 2 includes input means such as a monitor and a keyboard, and communication means for the network 4. The user terminal 2 accepts a search request for content by a user and sends a search term stored in the search request via the network 4. It transmits to the search word receiving part 31 of the information search device 3 mentioned later. Further, the user terminal 2 receives a search result transmitted from the search result output unit 35 of the information search device 3 described later via the network 4 and presents the search result to the user.

  The information search device 3 includes a search word receiving unit 31 that receives a search word for searching for content, a content search unit 32 that searches for content including the received search word, and a browsing history that searches the browsing history of the searched content A search unit 33, a fitness synthesis unit 34 that synthesizes the text suitability and browsing popularity of the searched content, and a search result output unit 35 that outputs a content search result based on the synthesized result. The information retrieval apparatus includes a storage unit described below.

  The content storage unit 301 is connected to the content search unit 32, and as shown in FIG. 2, content text information indicating the content itself, a content identifier that can identify the content, and content in which the content is stored Stored in association with the location. The content sentence information is a sentence indicating the content of the content. For example, if the content is a page stored on a server on the web, it means a sentence written on the page. The content identifier makes it possible to identify a certain content from other content. For example, if the content is a page stored in a server on the web, it means a URL (Uniform Resource Locator) of the page. The content location is a location where content files are stored. For example, when content is stored in a local computer, path information as shown in FIG. In the case of content, a URL is stored.

  The browsing history log 302 is configured in a file format as a log, is connected to the browsing history search unit 33, and stores browsing history of content browsed in the past by the user. For example, as shown in FIG. 3, a plurality of content identifiers are stored in a form corresponding to one browsing request per line, and the content “www.goo.ne.jp/” is browsed first, Next, it means that the content “www.nttr.co.jp/genki.html” has been browsed. The content identifier stored in the browsing history log 302 is updated separately from the information search system 1 every time a browsing request is received from an arbitrary user.

  As the network 4, for example, a LAN (Local Area Network), the Internet, a public line network, a cable television network, or the like can be used. The Internet is a collection of networks linked to each other based on a predetermined protocol.

  Next, processing of the information search device 3 in the embodiment of the present invention will be described with reference to FIG. FIG. 4 is a flowchart showing processing of the information search device 3 in the first embodiment.

  First, the search word receiving unit 31 receives a search word transmitted from the user terminal 2 via the network 4, and transmits the search word to the content search unit 32 (S101).

  Next, the content search unit 32 receives the search term transmitted from the search term receiving unit 31, reads the content and the content identifier from the content storage unit 301, a plurality of contents including the received search term, and the plurality of the content Search multiple content identifiers corresponding to each content, calculate the text suitability of each searched content, and browse the content search results that associate each searched content identifier with each calculated text suitability The data is transmitted to the search unit 33 and the fitness composition unit 34 (S102).

  Here, the sentence suitability means the frequency with which a search word appears in the content, and for example, TF (Term Frequency) method, TF / IDF (Term Frequency / Inverse Document Frequency) method, etc. are common. Has been used.

  Subsequently, the browsing history search unit 33 receives the content search result transmitted from the content search unit 32, reads the content identifier browsed in the past from the browsing history log 302, and the content identifier included in the received content search result A content identifier having the same content identifier is searched, a browsing popularity indicating the frequency of browsing the content of the content identifier included in the content search result is calculated, and transmitted to the matching level synthesizing unit 34 (S103).

  Here, the browsing popularity calculation method in the present embodiment will be described. When the content identifier same as the content identifier received from the content search unit 32 is stored in the browsing history log 302, the browsing popularity (S) is the number of rows of the content identifier stored in the browsing history log 302. The number of requests (C) can be calculated by, for example, Expression (1).

S = C Formula (1)
By calculating the browsing popularity using equation (1), the browsing popularity of the content that has been browsed a lot in the past will increase, and by combining with the score of the content to be described later, the popular content will be given priority Can be provided to users. In addition, it is also possible to use arbitrary functions, such as Formula (2)-Formula (4), instead of Formula (1).

S = log (C) (2)
S = exp (C) (3)
S = f (C) (4)
Thereafter, the matching level synthesis unit 34 receives the content search result transmitted from the content search unit 32 and the browsing popularity level transmitted from the browsing history search unit 33, and calculates the sentence matching level and the browsing popularity level for each content identifier. The combined score is calculated, and the score is transmitted to the search result output unit 35 (S104). The score calculation method may be calculated using addition, multiplication, or the like, or an arbitrary function may be used as in the case of browsing popularity.

  The processes of (S103) and (S104) are repeated for all content identifiers searched in (S102) (S105).

  The search result output unit 35 receives the score for each content identifier transmitted from the fitness synthesis unit 34 and outputs the search result in which the content identifiers are rearranged in the order of the highest score to the user terminal 2 via the network 4. (S106). In addition, for example, only the top N items (N is a natural number of 1 or more) among the content identifiers arranged in descending order of the scores can be output as the search results.

  Through the above processing, for example, a search result as shown in FIG. 5 can be presented to the user via the screen of the user terminal 2.

  Finally, the reason why the content that matches the search word and is popular is preferentially provided to the user by synthesizing the sentence matching degree and the browsing popularity degree will be described.

  FIG. 6 is an explanatory diagram for explaining the effect of the first embodiment. Here, it is assumed that the search term used by the user is “Tokyo”. The contents obtained as a search result are “Tokyo information”, “Tokyo Metropolitan Government”, “Tokyo tourism”, “University of Tokyo”, “Tokyo station”, “weather forecast”. Suppose you want search results for. Furthermore, it is assumed that the sentence suitability for each content is 147, 150, 152, 145, 142, 31 and the browsing popularity is 32, 27, 23, 21, 18, 100.

  Here, the text suitability is based on the frequency of occurrence of the character string “Tokyo” in the content. Therefore, for the content where the name of the tourist destination is often used such as “Sightseeing in Tokyo”. On the other hand, it is not directly related to the subject of the content such as “weather forecast”, but it is a small value for content that appears only once in the content.

  In the case of the classical method, as shown in FIG. 6 (a), the score is scored according to the sentence matching degree. Therefore, the search results are ordered in terms of only the number of appearances of “Tokyo” in the content, and in the order that the user seeks. Cannot be output.

  Further, in the case of the conventional method, as shown in FIG. 6B, the browsing popularity is scored, so that the search results can be provided in the order of contents popular to the user. However, on the other hand, content that is searched and viewed by various search terms such as “weather forecast” is presented at the top, so that such popular content is the search term given by the user. It is not necessarily content with a high degree of fitness.

  On the other hand, in the case of the method of synthesizing the sentence matching degree and the browsing popularity degree described in the present embodiment, it is possible to provide search results in the order desired by the user as shown in FIG. In addition, it is possible to rank so that popular contents are highly displayed and have a high degree of matching with a search term.

  According to the present embodiment, a content including a search word for searching for content and a content identifier corresponding to the content are searched from the content storage unit 301, and the frequency of appearance in the content searched by this search word is indicated. The sentence matching degree is calculated, and a content identifier having the same content identifier as the searched content identifier is searched from the browsing history log 302, and the browsing popularity degree indicating the frequency of browsing the content of the searched content identifier is calculated. Since the content identifier searched based on the result of combining the calculated sentence fitness and the browsing popularity is output, it is possible to display the content reflecting the user's preference.

  According to the present embodiment, the matching degree synthesis unit 34 calculates a score obtained by synthesizing the sentence matching degree and the browsing popularity degree, and the search result output unit 35 selects the content identifiers in descending order of the score. Therefore, it is possible to preferentially provide a popular content identifier to the user with a high degree of matching with the search term input by the user.

[Second Embodiment]
FIG. 7 is a configuration diagram showing the configuration of the information search system 1 in the second embodiment. The configuration of the information search system 1 in the present embodiment is basically the same as that of the first embodiment except for the browsing history storage unit 302, and therefore, a duplicate description is omitted here.

  The browsing history storage unit 302 according to the present embodiment is configured in a database format. As shown in FIG. 8, the browsing history storage unit 302 is a browsing that shows the content identifier of the browsed content and the browsing time when the content was browsed. The request date and time and the user identifier of the user who has browsed the content are stored in association with each other in a form corresponding to a single browsing request. In the case of FIG. 8, the content having the content identifier “www.goo.ne.jp/” at the time “12:32:54 on Feb. 27, 2007” is first used as “75023997047”. It means that it was browsed by a user with a user identifier. Further, the browsing request date and time, the content identifier, and the user identifier stored in the browsing history storage unit 302 are updated separately from the information search system 1 every time a browsing request is received from an arbitrary user.

  Next, processing of the information search device 3 in the embodiment of the present invention will be described with reference to FIG. FIG. 9 is a flowchart showing processing of the information search device 3 in the second embodiment.

  First, the search word receiving unit 31 receives a search word transmitted from the user terminal 2 via the network 4 and transmits the search word to the content search unit 32 (S201).

  Next, the content search unit 32 receives the search term transmitted from the search term receiving unit 31, reads the content and the content identifier from the content storage unit 301, a plurality of contents including the received search term, and the plurality of the content Search multiple content identifiers corresponding to each content, calculate the text suitability of each searched content, and browse the content search results that associate each searched content identifier with each calculated text suitability The data is transmitted to the search unit 33 and the fitness composition unit 34 (S202). Note that the method for calculating the sentence matching degree is the same as that in the first embodiment.

  Subsequently, the browsing history search unit 33 receives the content search result transmitted from the search word receiving unit 31, and reads the content identifier, the browsing request date and time, and the user identifier that were browsed in the past from the browsing history storage unit 302. And searching for a content identifier having the same content identifier as the content identifier included in the received content search result, calculating a browsing popularity indicating the frequency of browsing the content with the content identifier included in the content search result, The data is transmitted to the combining unit 34 (S203).

  Here, the browsing popularity calculation method in the present embodiment described in (S203) will be specifically described. FIG. 10 is a flowchart illustrating a process of calculating the browsing popularity according to the second embodiment.

  First, a value of 0 (zero) is set to the variable a used for calculating the browsing popularity (S203-1).

  Next, the current time is acquired (S203-2).

  Subsequently, the content identifier browsed in the past, the browse request date and time, and the user identifier are read from the browse history storage unit 302, and a content identifier having the same content identifier as the content identifier included in the received content search result is searched. The browsing request date and time and the user identifier corresponding to the searched content identifier are acquired. Then, for each acquired user identifier, only the one with the newest browsing request date and time is left, and a list showing the relationship between the user identifier and the final browsing request date and time shown in FIG. 11 is created (S203-3).

  After that, the difference between the current time acquired in (S203-2) and the last browsing request date and time of the list created in (S203-3) is calculated, and the time from when each user last viewed the content to the present Time E is calculated (S203-4).

  Then, the calculated fresh time S is calculated by substituting the calculated elapsed time E into the equation (5), and is added to the browsing popularity a (S203-5).

S = exp (−E) (5)
The browsing freshness degree obtained by the equation (5) becomes smaller as the elapsed time E from the last browsing time to the present time becomes longer, and becomes larger as the elapsed time E becomes shorter. Therefore, in the case of popular content that is browsed frequently, the elapsed time E generally tends to be small, so that the value of the browsing popularity increases. Therefore, by adding this browsing freshness to the browsing popularity, it is possible to obtain the browsing popularity based on the browsing frequency of the user. In addition, it is also possible to use arbitrary functions, such as Formula (6)-Formula (10), instead of Formula (5).

S = 1 / E (6)
S = 1 / log (E) (7)
S = c × exp (−bE /) (8)
S = c / bE Formula (9)
S = c / log (bE) (10)
However, c and b are constants.

  The processes of (S203-4) and (S203-5) are repeated for all user identifiers searched in (S203-3) (S203-6).

  The value of the variable a finally obtained is used as the browsing popularity of the content, and is transmitted to the fitness composition unit 34 (S203-7). Note that the value calculated as log (a) can be used as the browsing popularity instead of the value of the variable a as it is.

  Returning to FIG. 9, the matching level synthesis unit 34 receives the content search result transmitted from the content search unit 32 and the browsing popularity level transmitted from the browsing history search unit 33, and the sentence matching level and browsing for each content identifier. A score obtained by combining the popularity is calculated, and the score is transmitted to the search result output unit 35 (S204).

  The processes of (S203) and (S204) are repeated for all content identifiers searched in (S202) (S205).

  The search result output unit 35 receives the score for each content identifier transmitted from the fitness synthesis unit 34 and outputs the search result in which the content identifiers are rearranged in the order of the highest score to the user terminal 2 via the network 4. (S206).

  Finally, the reason for not being influenced by the intensive browsing of some users by calculating the browsing popularity considering the browsing freshness will be described.

  FIG. 12 is an explanatory diagram for explaining the effect of the second embodiment. Here, it is assumed that content A and content B are included in a search result for a certain search request. Further, the past browsing history searched by the browsing history search unit 33 in the browsing history storage unit 302 has a total browsing number of 17 for both the content A and the content B as shown in FIG. 12 (FIG. 12). It is assumed that there is a large difference in the viewing status of each user.

  In the case of the content A, it is viewed continuously by a specific user 2, but since it is hardly viewed by other users, it is considered to be an unpopular content, but due to mischief of some users It can be considered that browsing was done. On the other hand, in the case of the content B, since a plurality of users are viewed relatively frequently with the same frequency, the content B is a popular content and is considered to be frequently viewed by many users.

  Here, when the browsing popularity of the contents A and B is calculated by the number of browsing as in the conventional method, the browsing popularity of both is the same at 17, and the browsing popularity is the original popularity. Not reflected. On the other hand, according to the calculation of the browsing popularity according to the present embodiment, as shown in FIG. 12, the content A is 0.90 and the content B is 1.27. It becomes larger and it is possible to obtain the browsing popularity reflecting the original popularity.

  Therefore, the calculation of the browsing popularity in this embodiment can obtain a popularity that faithfully reflects the actual popularity without being affected by the intensive browsing of some users. By presenting the search results ranked based on the scores with added degrees to the user, the user can efficiently search for popular content.

  Subsequently, the reason why the calculation of the browsing popularity according to the present embodiment is not affected by the temporary intensive browsing will be described.

  FIG. 13 is an explanatory diagram for explaining the effect of the second embodiment. Assume that content A and content B are included in a search result for a certain search request, as in the case described with reference to FIG. In addition, as shown in FIG. 13, the past browsing history searched by the browsing history search unit 33 in the browsing history storage unit 302 has a total browsing count of 17 for both the content A and the content B. Suppose that there is a big difference in the browsing situation of the person.

  In the case of the content A, the browsing time is limited to a certain period in the past, and since it has not been browsed recently, it is considered to be an unpopular content. This means that a lot of browsing has been done. On the other hand, in the case of the content B, since a plurality of users browse relatively frequently with the same frequency, it is a content that is permanently more popular than the content A, and many users frequently view it. It means that it has continued.

  Here, when the browsing popularity of the contents A and B is calculated by the number of browsing as in the conventional method, the browsing popularity of both is the same at 17, and the browsing popularity is the original popularity. Not reflected. On the other hand, according to the calculation of the browsing popularity according to the present embodiment, as shown in FIG. 13, the content A is 0.03 and the content B is 1.27. It becomes larger and it is possible to obtain the browsing popularity reflecting the original popularity.

  Therefore, the calculation of the browsing popularity in the present embodiment can obtain a popularity that more faithfully reflects the actual popularity without being affected by the temporary intensive browsing. By presenting the search results ranked based on the score with the added to the user, the user can efficiently search for popular content.

  According to the present embodiment, since the browsing freshness indicating the elapsed time from the time when the same content identifier as the content identifier searched by the search word is browsed to the present is added to the browsing popularity, the original popularity degree Can be provided to the user.

  According to the present embodiment, since the freshness of browsing is calculated using the newest browsing time, it is possible to provide the user with a content identifier that more reliably reflects the original popularity.

[Third Embodiment]
FIG. 14 is a configuration diagram illustrating a configuration of the information search system 1 according to the third embodiment. The configuration of the information search system 1 in the present embodiment is basically the same as that of the first embodiment except for the browsing history storage unit 302 and the user attribute storage unit 303, and therefore, a duplicate description is omitted here. To do.

  The browsing history storage unit 302 in the present embodiment is configured in a database format, and as shown in FIG. 15, a browsing that indicates a content identifier of the browsed content and a browsing time when the content is browsed. The request date and time, the user identifier of the user who browsed the content, and the browse time when the content was browsed are stored in association with each other in a form corresponding to one browse request. This browsing time is the length from the time when browsing is started to the time when browsing of another content is started or when communication is interrupted. Further, when the browsing time exceeds a certain value, it is possible that the user has left the terminal and an appropriate value can be set as the browsing time.

  The user attribute storage unit 303 is configured in a database format, and stores the user identifier of the user and the attribute of the user in association with each other as shown in FIG. The attribute is specific information about the user such as sex, age, hobby, occupation, nationality, address, and the like.

  Next, processing of the information search device 3 in the embodiment of the present invention will be described with reference to FIG. FIG. 17 is a flowchart showing the processing of the information search device 3 in the third embodiment.

  First, the search word receiving unit 31 receives the search word and the user attribute transmitted from the user terminal 2 via the network 4, and transmits the search word to the content search unit 32 (S301). The user attributes may be input by the user when the user issues a search request, or attribute information set in advance in the user terminal 2 is automatically transmitted. Also good.

  Next, the content search unit 32 receives the search term and the user attribute transmitted from the search term receiving unit 31, reads the content and content identifier from the content storage unit 301, and includes a plurality of received search terms. The content and the plurality of content identifiers corresponding to the plurality of contents are searched, the sentence suitability of each searched content is calculated, and each searched content identifier is associated with each calculated sentence suitability The content search result is transmitted to the browsing history search unit 33 and the matching level synthesis unit 34, and the user attributes are also transmitted to the browsing history search unit 33 and the matching level synthesis unit 34 (S302). Note that the method for calculating the sentence matching degree is the same as that in the first embodiment.

  Subsequently, the browsing history search unit 33 receives the content search result and the user attribute transmitted from the search word receiving unit 31, and the content identifier, the browsing request date and time, and the browsing request date and time used in the past from the browsing history storage unit 302. The content identifier having the same content identifier as the content identifier included in the received content search result is retrieved, and the browsing popularity indicating the frequency of browsing the content with the content identifier included in the content search result is obtained. It calculates and transmits to the fitness synthesis unit 34 (S303).

  Here, the method for calculating the browsing popularity in the present embodiment described in (S303) will be specifically described. FIG. 18 is a flowchart illustrating a process of calculating the browsing popularity according to the third embodiment.

  First, a value of 0 (zero) is set to the variable a used for calculating the browsing popularity (S303-1).

  Next, the current time is acquired (S303-2).

  Subsequently, the content identifier browsed in the past, the browse request date and time, and the user identifier are read from the browse history storage unit 302, and a content identifier having the same content identifier as the content identifier included in the received content search result is searched. The browsing request date and time and the user identifier corresponding to the searched content identifier are acquired. Then, for each acquired user identifier, only the one with the latest browsing request date and time is left, and the list A showing the relationship between the user identifier and the final browsing request date and time shown in FIG. 11 described in the second embodiment. Create In addition, the list A is created, and the user identifier and the user attribute are read from the user attribute storage unit 303, and the user identifier having the same attribute as the received user attribute is searched and searched. A list B indicating the relationship between user identifiers and user attributes is created. Thereafter, the user identifiers not included in the list B and the browsing time corresponding to the user identifiers are deleted from the list A (S303-3).

  The search conditions for user attributes at the time of creating List B can be any combination such as “sex only”, “sex age”, “age difference within 3 years”, “age and nationality”, etc. Can be set.

  As a result, when searching the browsing history of content, the browsing popularity is calculated using only the past browsing history by the user similar to the user's attribute, so it is popular with users close to the user's attribute It becomes possible to present certain content with priority.

  When creating the list B, the number of hits of users having the same attributes as the received user attributes, that is, the total number of user identifiers of the list A after deletion is greater than a predetermined threshold value H set in advance. If there are many, the processing of (S303-6) described later is started, and if it is less than the predetermined threshold H, the processing of (S304-5) is started (S304-4). Note that, for example, an arbitrary value such as 20 can be set as the predetermined threshold.

  A user identifier corresponding to a content identifier that matches a part of the content identifier included in the content search result is read from the browsing history storage unit 302, and the read user identifier and the browsing time corresponding to this user identifier are listed. A is added to A (S304-5).

  For example, when the content identifier included in the content search result is “http://www.goo.ne.jp/˜tokyo/gaiden/ushi/gus.html”, the domain name “www.goo. “ne.jp” is extracted, and a content identifier having the same domain name as this domain name as a content identifier is searched from the browsing history storage unit 302.

  It is also possible to use a combination of arbitrary elements separated by symbols such as “/”, “.”, “˜” instead of the domain name as a search condition. For example, “www.goo.ne.jp/˜tokyo”, “www.goo.ne.jp/˜tokyo/gaiden”, “www.goo.ne.jp/˜tokyo/gaiden/ushi”, “goo. A part of the search condition may be “ne.jp”, “tokyo”, “gus.html”, or the like.

  As a result, content with a rare content in the past browsing history can also be searched for content having a content identifier element common to the content identifier included in the content search result, so that more related browsing search information can be obtained. Obtainable. For example, even if the content has no browsing history in the past, if there is a browsing history for content with the same domain name as that content, the browsing popularity of the content can be calculated.

  After that, the difference between the current time acquired in (S303-2) and the last browse request date / time of list A created in (S303-3) or (S303-5) is calculated, and each user puts the content at the end. Elapsed time E from the point of view to the present is calculated (S303-6).

  Then, the browsing time for each user identifier is read from the browsing history storage unit 302, and the browsing freshness S is calculated by substituting the read browsing time D and the calculated elapsed time E into the equation (11). Add to degree a (S303-7).

S = log (D) × exp (−E) (11)
As a result, in addition to the popularity reflecting the frequency of browsing by the user, the browsing popularity increases as the browsing time when the user browses the content, and conversely, the browsing popularity decreases as the browsing time decreases. Since the degree becomes small, the user can efficiently search for popular contents by ranking the search results according to the scores added with the browsing popularity and presenting them to the user. In addition, it is also possible to use arbitrary functions, such as Formula (12)-Formula (15), instead of Formula (11).

S = D × exp (−E) (12)
S = c × D × exp (−E) (13)
S = log (c × D) × exp (−E) (14)
S = f (D) × f (E) (15)
However, c is a constant.

  Note that the browsing freshness S may be calculated using only the elapsed time E without using the browsing time D, as described in the second embodiment.

  The processes of (S303-6) and (S303-7) are repeated for all user identifiers searched in (S303-3) or (S303-5) (S303-8).

  The value of the variable a finally obtained is used as the browsing popularity of the content, and is transmitted to the fitness composition unit 34 (S303-9). Note that the value calculated as log (a) can be used as the browsing popularity instead of the value of the variable a as it is.

  Returning to FIG. 17, the matching level synthesis unit 34 receives the content search result transmitted from the content search unit 32 and the browsing popularity level transmitted from the browsing history search unit 33, and the sentence matching level and browsing for each content identifier. The score obtained by combining the popularity is calculated, and the score is transmitted to the search result output unit 35 (S304).

  The processes of (S303) and (S304) are repeated for all content identifiers searched in (S302) (S305).

  The search result output unit 35 receives the score for each content identifier transmitted from the fitness synthesis unit 34 and outputs the search result in which the content identifiers are rearranged in the order of the highest score to the user terminal 2 via the network 4. (S306).

  According to the present embodiment, a user identifier having the same attribute as the received user attribute is retrieved from the user attribute storage unit 303, and a list B indicating the relationship between the retrieved user identifier and the attribute is obtained. Since the freshness of browsing is calculated using the list A in which the created user A and the user identifiers not stored in the list B and the browsing time corresponding to the user identifier are deleted, the freshness of browsing is calculated. It is possible to preferentially present content that is popular with users close to the attribute.

  According to the present embodiment, when the total number of user identifiers in the list A is smaller than a predetermined threshold, the user identifier corresponding to the content identifier that matches a part of the content identifier searched by the content search means is Reading from the browsing history storage unit 302 and adding the read user identifier and the browsing time corresponding to this user identifier to the list A, so that the browsing freshness is calculated using more related browsing information Can do.

  According to the present embodiment, the browsing time corresponding to each user identifier in the list A is read from the browsing history storage unit 302, and the browsing freshness is calculated by further using this browsing time. Can be provided to the user.

[Fourth Embodiment]
In the fourth embodiment, the search result is presented to the user according to the first to third embodiments, and the content identifier output as the search result is selected by the user, and then the selected content identifier is selected. The process until the content is presented to the user will be described.

  FIG. 19 is a configuration diagram illustrating a configuration of the information search system 1 according to the fourth embodiment. The information search system 1 according to the present embodiment includes a user terminal 2, an information search device 3, and a web server 5, which are connected to each other via a network 4.

  The information search device 3 is analyzed by a content identifier receiving unit 36 that receives a content identifier, a content acquisition unit 37 that acquires the content of the content identifier, and a content analysis unit 38 that analyzes related content of the acquired content. An information search apparatus further comprising a browsing history search unit 39 that searches a browsing history of related content, a content processing unit 40 that processes content based on an analysis result, and a content output unit 41 that outputs the processed content. Thus, the browsing history storage unit 302 described in the first to third embodiments is provided.

  The function of the browsing history search unit 39 described in the present embodiment is the same function as the browsing history search unit 33 described in the first to third embodiments, and the browsing history search unit 33 searches the browsing history. It is also possible to operate as the unit 39.

  Next, processing of the information search device 3 in the embodiment of the present invention will be described with reference to FIG. FIG. 20 is a flowchart showing the processing of the information search apparatus 3 in the fourth embodiment.

  First, when a certain content identifier is selected from among a plurality of content identifiers presented to the user as a search result, the content identifier receiving unit 36 receives the selected content identifier via the network 4. And transmitted to the content acquisition unit 37 (S401).

  Next, the content acquisition unit 37 receives the content identifier transmitted from the content identifier reception unit 36, acquires the content having the content identifier from the web server 5, and transmits it to the content analysis unit 38 and the content processing unit 40. (S402).

  Subsequently, the content analysis unit 38 receives the content transmitted from the content acquisition unit 37, analyzes the related content in the content directly related to the content, and extracts the related content identifier of the analyzed related content. Then, the extracted related content identifier is transmitted to the browsing history search unit 39 (S403).

  Here, for example, if the transmitted content is a web page, the related content means a hyperlink destination content in the web page, and the related content identifier is a URL of the related content. means.

  Thereafter, the browsing history search unit 39 receives the related content identifier transmitted from the content analysis unit 38, reads the content identifier browsed in the past from the browsing history storage unit 302, and is the same as the related content identifier in the received content. A content identifier is searched, a browsing popularity degree indicating the frequency of browsing the content of the related content identifier is calculated, and transmitted to the content processing unit 40 (S404). Note that the browsing popularity calculation method according to the present embodiment may be any of the calculation methods described in the first to third embodiments.

  The content processing unit 40 receives the content transmitted from the content acquisition unit 37 and the browsing popularity of the related content identifier transmitted from the browsing history search unit 39, and processes the content based on the browsing popularity value. (S405).

  For example, when the browsing popularity value is high, the character size of the anchor text of the corresponding hyperlink is processed to be larger, and when the browsing popularity value is low, it is processed to be smaller. The degree of the size may be changed so as to be proportional to the value of the browsing popularity, or a difference from a predetermined value is calculated, and when the difference is positive, the size is reduced by 15 points, minus In this case, 10 points may be used. In the case of FIG. 21, the browsing popularity of the link destination content of the hyperlink corresponding to the anchor texts of “language” and “next” is large, while the browsing popularity of “human” is small.

  In this way, by changing the display size of the anchor text according to the level of browsing popularity, the user can search for popular content more easily.

  The processes of (S404) and (S405) are repeated for all the related content identifiers searched in (S403) (S406).

  Finally, the content output unit 41 outputs the content transmitted from the content processing unit 40 to the user terminal 2 via the network 4 (S407).

  As described above, the case where the content identifier output as the search result is selected by the user and then the content is presented to the user has been described. However, in the present embodiment, the content output as the search result is selected. However, the present invention is not limited to this, and the same application may be applied to a case where browsing is reached through an arbitrary route, and even in that case, the same effect as described here can be obtained.

  According to the present embodiment, for the content with the content identifier selected by the user, the related content in the content directly related to the content is analyzed, the browsing popularity for the analyzed related content is calculated, Since the content processed based on the browsing popularity is output, it is possible to easily search for related content that is popular in the output content.

[Fifth Embodiment]
FIG. 22 is a configuration diagram showing the configuration of the information search system 1 in the fifth embodiment. The information search system 1 in the present embodiment has a configuration further including the user attribute storage unit 303 described in the third embodiment in addition to the configuration described in the fourth embodiment.

  In addition to the configuration described in the third embodiment, the browsing history storage unit 302 described in the present embodiment includes content identifiers stored in the browsing history storage unit 302 as shown in FIG. The transition source content identifier of the transition source content that was browsed before the content was browsed is further stored. For example, when browsing the content A while browsing the content A while tracing the hyperlink displayed on the content A, the content A is the transition source content, and the content identifier of the content A is the transition source content identifier. Become.

  The transition source content can be easily obtained from a record in which all the content viewed by the user is arranged in time order. For example, in the case of a mobile phone operator, all browsing request information of the user can be obtained from the log of the communication device, so that information on the transition source content can also be obtained. Moreover, even if the transition is not based on a hyperlink, the content that was browsed most recently in time may be used as the transition source content, and when the user makes a browsing request by directly entering the content identifier. May have no transition source content.

  Next, processing of the information search device 3 in the embodiment of the present invention will be described with reference to FIG. FIG. 24 is a flowchart illustrating the processing of the information search device 3 in the fifth embodiment.

  First, when a certain content identifier is selected from among a plurality of content identifiers presented to the user as a search result, the content identifier receiving unit 36 receives the selected content identifier via the network 4. And transmitted to the content acquisition unit 37 (S501).

  Next, the content acquisition unit 37 receives the content identifier transmitted from the content identifier reception unit 36, acquires the content having the content identifier from the web server 5, and transmits it to the content analysis unit 38 and the content processing unit 40. At the same time, the received content identifier is transmitted to the content analysis unit 38 (S502).

  Subsequently, the content analysis unit 38 receives the content and the content identifier transmitted from the content acquisition unit 37, analyzes the related content in the content directly related to the content, and relates the related content in the analyzed content. The content identifier is extracted, and the related content identifier in the extracted content and the received content identifier are transmitted to the browsing history search unit 39 (S503).

  Next, the browsing history search unit 39 receives the related content identifier and the content identifier in the content transmitted from the content analysis unit 38, analyzes and analyzes the related content outside the content indirectly related to the content. The related content identifier of the related content outside the content is extracted, and the extracted related content identifier outside the content and the importance with respect to the related content identifier outside the content are transmitted to the content processing unit 40 (S504).

  Regarding the extraction method of the related content identifier outside the content and the calculation method of the importance in (S504), (i) the content having the same content identifier and the transition source content identifier is extracted as “related content outside the content”. And (ii) searching the browsing history for content having the content of the received content identifier as the source content, and searching for content not included in the hyperlink of the content of the received content identifier from the related content outside the content "And the method of extraction as follows. The received content identifier will be described below as “URL0”.

  First, the method (i) will be described.

  The browsing history search unit 39 reads the content identifier from the browsing history storage unit 302, searches for a content identifier having the same content identifier as the received content identifier, and displays the transition source content identifier corresponding to the searched content identifier as the browsing history. A list C of transition source content identifiers that is read from the storage unit 302 and associates the transition source content identifiers with the number of the transition source content identifiers is created (S504-1a).

  For example, in the case of the transition source content identifier list C shown in FIG. 25A, the number of rows of the transition source content identifier “URL1” is five for the received content identifier “URL0”, and “URL2” This means that the number of lines is seven, and the number of lines of “URL3” is two.

  Further, when the transition source content identifier is read from the browsing history storage unit 302 in (S504-1a), further conditions can be given. For example, the browsing history search unit 39 reads the user identifier and the user attribute from the user attribute storage unit 303 and searches for a user identifier having the same attribute as the user attribute received from the user terminal 2. Then, a list D indicating the relationship between the searched user identifier and the user attribute is created. Then, the transition source content identifier having the same user identifier as the user identifier included in the list D is read from the browsing history storage unit 302 to create the list C.

  It should be noted that whether or not the list D can be used can also be determined according to the search result of the user identifier. In addition, the user attributes used when creating the list D may be input by the user when the user issues a search request in (S501), or set in advance in the user terminal 2. The attribute information may be automatically transmitted. Furthermore, the user identifier used in the third embodiment may be used.

  Next, for each transition source content identifier included in the list C of transition source content identifiers created in (S504-1a), a transition source content identifier having the same transition source content identifier as the transition source content identifier is stored in the browsing history. A list E of transition destination content identifiers in which the content identifier corresponding to the searched transition source content identifier (hereinafter referred to as “transition destination content identifier”) and the number of the transition destination content identifiers are associated with each other is retrieved from the section 302. Create (S504-2a).

  For example, in the case of FIG. 25B, the number of lines of the transition destination content identifier “URL11” is 3 and the number of lines of “URL12” is 6 with respect to the transition source content identifier “URL1”. I mean.

  Subsequently, the importance of each transition destination content identifier included in the list E is calculated using the list C and the list E (S504-3a).

  Specifically, as illustrated in FIG. 25C, for example, for the transition destination content “URL11”, 15 obtained by multiplying 5 of the transition source content “URL1” and 3 of the transition destination content “URL11”; 14 which is obtained by multiplying 7 of the transition source content “URL2” and 2 of the transition destination content “URL11” is added, and the added total value 29 is set as the importance. Note that the importance G can be calculated by using the equations (16) to (20) instead of multiplying the number of rows.

G = log (C) × log (E) (16)
G = C × log (E) (17)
G = log (C) × E (18)
G = log (C) + log (E) Expression (19)
G = log (C) + E (20)
Note that C is the number of rows of the transition source content identifier, and E is the number of rows of the transition destination content identifier.

  Then, the browsing history search unit 39 sets each transition destination content identifier as a related content identifier outside the content, and transmits the related content identifier outside the content and the importance to the content processing unit 40 (S504-4a).

  Next, the method (ii) will be described.

  The browsing history search unit 39 reads the content identifier from the browsing history storage unit 302, searches for a content identifier having the same content identifier as the received content identifier, and displays the transition source content identifier corresponding to the searched content identifier as the browsing history. As shown in FIG. 26A, the list C that is read from the storage unit 302 and associates the transition source content identifier with the number of the transition source content identifiers is created as the transition destination content identifier (S504-1b). It is also possible to give further conditions when reading the transition source content identifier from the browsing history storage unit 302 using the list D used in the method described in (i).

  Next, as shown in FIG. 26B, a list E of related content identifiers in the content extracted in (S503) is created, and as shown in FIG. A list F not included in the list is created (S504-2b).

  Then, the browsing history search unit 39 sets each transition destination content identifier in the list F as a related content identifier outside the content, and uses the number of each transition destination content identifier as the importance, and the related content identifier outside the content and the important content identifier. The degree is transmitted to the content processing unit 40 (S504-3b).

  Returning to FIG. 24, the browsing history search unit 39 reads the content identifier browsed in the past from the browsing history storage unit 302 and searches for the same content identifier as the related content identifier in the received content together with the processing of (S504). Then, the browsing popularity corresponding to the frequency of browsing the content of the related content identifier is calculated and transmitted to the content processing unit 40 (S505). As in the fourth embodiment, the calculation method of the browsing popularity in this embodiment may be any of the calculation methods described in the first to third embodiments.

  The content processing unit 40 receives the content transmitted from the content acquisition unit 37, the browsing popularity for the related content identifier in the content transmitted from the browsing history search unit 39, the related content identifier outside the content, and its importance. The content is processed based on the browsing popularity value, and the related content identifier outside the content is processed according to the importance (S506).

  In the case of FIG. 27, the browsing popularity of the link destination content of the hyperlink corresponding to the anchor text of “language” and “next” is large, while the browsing popularity of “human” is small. Also, in the lower part of this content, the related content identifier and the importance level of the related content outside the content are displayed in a separately provided space. The related content identifier inside / outside the content may be a link to the content as a hyperlink. Further, instead of using the anchor text as the related content identifier itself, a title character string or the like of the link destination content may be displayed. Further, the display attribute of each hyperlink may be changed according to the importance value. For example, the hyperlink of related content with high importance may be large and the hyperlink of related content with low importance may be displayed small. Further, the display order of hyperlinks may be arranged according to the degree of importance. For example, hyperlinks may be arranged in descending order of importance. Furthermore, the location where the related content identifier outside the content is displayed may be on the right side, the left side, or the upper side of the content, or may be displayed in a separate window.

  As a result, when content is browsed in the order of content A → content B → content A → content C, content C is displayed as related content outside the content to the user browsing the content B. The content C can be browsed without going through the content A.

  In addition, when browsing in the order of content A → content B, even if a hyperlink to content B in content A is deleted at some time, content B is presented as related content outside the content. The user can browse the content B that does not exist in the hyperlink in the content A.

  Furthermore, since the importance of the browsing content outside the content can be grasped by the display attribute and display order, the popular content can be browsed more efficiently.

  Finally, the processing described in the first to fifth embodiments is executed by a program, and the program can be stored in a recording medium so as to be reproducible.

  According to the present embodiment, since the related content identifier outside the content indirectly related to the content is further output, the convenience of the content can be improved.

  According to the present embodiment, related content outside the content is output according to the importance and / or together with the importance, so that popular content can be browsed more efficiently.

It is a block diagram which shows the structure of the information search system in 1st Embodiment. It is a figure which shows the table stored in the content storage part. It is a figure which shows the browsing history stored in the browsing history storage part in 1st Embodiment. It is a flowchart which shows the process of the information search device in 1st Embodiment. It is a figure which shows an example of the search result in 1st Embodiment. It is explanatory drawing explaining the effect in 1st Embodiment. It is a block diagram which shows the structure of the information search system in 2nd Embodiment. It is a figure which shows the browsing history stored in the browsing history storage part in 2nd Embodiment. It is a flowchart which shows the process of the information search device in 2nd Embodiment. It is a flowchart which shows the process which calculates the browsing popularity in 2nd Embodiment. It is a list | wrist which shows the relationship between a user identifier and the last browsing request date. It is explanatory drawing explaining the effect in 2nd Embodiment. It is explanatory drawing explaining the effect in 2nd Embodiment. It is a block diagram which shows the structure of the information search system in 3rd Embodiment. It is a figure which shows the browsing history stored in the browsing history storage part in 2nd Embodiment. It is a figure which shows the table stored in the user attribute storage part. It is a flowchart which shows the process of the information search device 3 in 3rd Embodiment. It is a flowchart which shows the process which calculates the browsing popularity in 3rd Embodiment. It is a block diagram which shows the structure of the information search system in 4th Embodiment. It is a flowchart which shows the process of the information search device in 4th Embodiment. It is a figure which shows an example of the content before and after a process in 4th Embodiment. It is a block diagram which shows the structure of the information search system 1 in 5th Embodiment. It is a figure which shows the browsing history stored in the browsing history storage part in 5th Embodiment. It is a flowchart which shows the process of the information search device 3 in 5th Embodiment. It is a figure which shows the flow which extracts the related content outside a content in the method of (i). It is a figure which shows the flow which extracts the related content outside the content in the method of (ii). It is a figure which shows an example of the content before the process in the 5th Embodiment, and after a process.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Information search system 2 ... User terminal 3 ... Information search device 4 ... Network 5 ... Web server 31 ... Search word receiving part 32 ... Content search part 33 ... Browsing history search part 34 ... Conformity synthesis part 35 ... Search result output part 36 ... Content identifier receiving unit 37 ... Content acquisition unit 38 ... Content analysis unit 39 ... Browsing history search unit 40 ... Content processing unit 41 ... Content output unit 301 ... Content storage unit 302 ... Browsing history log, browsing history storage unit 303 ... Usage Person attribute storage

Claims (26)

Content storage means for storing the content and a first content identifier that can identify the content in association with each other;
Browsing history storage means for storing a second content identifier of the browsed content;
A search word receiving means for receiving a search word for searching for content;
The content and the first content identifier are read from the content storage means, the content including the search word and the first content identifier corresponding to the content are searched, and the frequency at which the search word appears in the content is determined. A content search means for calculating the sentence suitability shown,
The second content identifier is read from the browsing history storage means, a second content identifier having the same content identifier as the searched first content identifier is searched, and the frequency at which the content of the first content identifier is browsed is determined. A first browsing history search means for calculating the browsing popularity shown;
Relevance synthesis means for synthesizing the sentence relevance and the browsing popularity;
Search result output means for outputting the searched first content identifier based on the combined result;
An information retrieval apparatus comprising: The browsing history storage means stores the second content identifier in units of rows,
The information search device according to claim 1, wherein the browsing popularity is calculated using the number of rows of the second content identifier having the same content identifier as the searched first content identifier. The fitness synthesis means calculates a score obtained by synthesizing the text fitness and the browsing popularity,
The information search device according to claim 1 or 2, wherein the search result output means outputs the first content identifiers in the order of the highest score. The browsing history storage means stores the second content identifier in association with the browsing time when the content of the second content identifier was browsed,
The first browsing history search means includes:
After searching for the second content identifier, calculation means for reading the browsing time corresponding to the second content identifier from the browsing history storage means and calculating a freshness of browsing indicating an elapsed time from the browsing time to the present When,
Adding means for adding the browsing freshness to the browsing popularity;
The information search apparatus according to claim 1, further comprising: The browsing history storage means stores the second content identifier, the browsing time when the content of the second content identifier was browsed, and the user identifier of the user who browsed the content in association with each other. ,
The first browsing history search means includes:
After searching for the second content identifier, the browsing time and the user identifier corresponding to the second content identifier are read from the browsing history storage means, and a list showing the relationship between the browsing time and the user identifier Creating means for creating A;
For each user identifier in the list A, calculating means for calculating a browsing freshness indicating an elapsed time from the browsing time to the present time;
Summing up the browsing freshness for all user identifiers in the list A, and adding means for adding the total browsing freshness to the browsing popularity,
The information search apparatus according to claim 1, further comprising:   The said calculation means reads the newest browsing time from the said browsing history storage means among the browsing times corresponding to the said 2nd content identifier, and calculates the said browsing freshness degree. The information retrieval device described.   7. The information search apparatus according to claim 4, wherein the browsing freshness value is a smaller value as the elapsed time is longer, and is a larger value as the elapsed time is shorter. . A user attribute storage means for storing the user identifier of the user and the attribute of the user in association with each other;
The browsing history search unit
When a user attribute is received, the user identifier and the attribute are read from the user attribute storage means, and a user identifier having the same attribute as the received attribute is searched, and the searched use Creating means for creating a list B indicating the relationship between the person identifier and the attribute;
Delete means for deleting a user identifier not stored in the list B from the list A and a browsing time corresponding to the user identifier;
The information search device according to claim 5, wherein the calculation unit calculates the browsing freshness using the list A after deletion.   The first browsing history search means corresponds to a content identifier that matches a part of the searched first content identifier when the total number of user identifiers in the list A after deletion is less than a predetermined threshold. 9. The information search apparatus according to claim 8, wherein a user identifier is read from the browsing history storage means, and the read user identifier and a browsing time corresponding to the user identifier are added to the list A. . The browsing history storage means further stores a browsing time during which the content of the second content identifier was browsed,
The said calculating means reads the browsing time corresponding to each user identifier of the said list | wrist from the said browsing history storage means, and calculates the said browsing freshness further using the said browsing time. Or the information search device of 9. Content identifier receiving means for receiving the first content identifier when the output first content identifier is selected by a user;
Content acquisition means for acquiring the content of the first content identifier;
A content analysis unit that analyzes the related content in the content directly related to the content from the acquired content and extracts the related content identifier of the related content;
The second content identifier is read from the browsing history storage means, the second content identifier having the same content identifier as the related content identifier is searched, and the browsing popularity indicating the frequency of browsing the content of the related content identifier is obtained. A second browsing history search means for calculating;
Content processing means for processing the content based on the browsing popularity;
Content output means for outputting the processed content;
The information search apparatus according to claim 1, further comprising: The second browsing history search means calculates the browsing popularity and extracts a related content identifier of related content outside the content indirectly related to the content,
The information search apparatus according to claim 11, wherein the content output means further outputs a related content identifier outside the content. The browsing history storage unit further stores a transition source content identifier of the transition source content that was browsed before the content of the second content identifier stored in the browsing history storage unit was browsed. ,
The second browsing history search means includes:
The second content identifier is read from the browsing history storage means, a content identifier having the same content identifier as the received first content identifier is searched, the transition source content identifier corresponding to the content identifier and the transition source content identifier Create a transition source content identifier list that correlates with the number of
A transition source content identifier having the same transition source content identifier as the transition source content identifier in the transition source content identifier list is searched from the browsing history storage means, and the searched transition source content identifier and the number of the transition source content identifiers are obtained. Create an associated transition destination content identifier list,
Using the transition source content identifier list and the transition destination content identifier list, the importance of the transition destination content identifier is calculated,
The information search apparatus according to claim 12, wherein the transition destination content identifier is extracted as a related content identifier outside the content. The browsing history storage unit further stores a transition source content identifier of a transition source content that was browsed before the content of the second content identifier stored in the browsing history storage unit was browsed. ,
The second browsing history search means includes:
The second content identifier is read from the browsing history storage means, a content identifier having the same content identifier as the received first content identifier is searched, the transition source content identifier corresponding to the content identifier and the transition source content identifier Create a list that associates the number of
Delete the related content identifier in the content from the transition destination content identifier list,
13. The information search according to claim 12, wherein the transition destination content identifier in the transition destination content identifier list after deletion is extracted as a related content identifier outside the content, and the number of the transition destination content identifiers is set as the importance. apparatus.   The information search device according to claim 13 or 14, wherein the content output means outputs a related content identifier outside the content according to the importance and / or together with the importance. Browsing history storage means for storing a second content identifier of the browsed content;
Content identifier receiving means for receiving a first content identifier for desired content;
Content acquisition means for acquiring the content of the first content identifier;
A content analysis unit that analyzes the related content in the content directly related to the content from the acquired content and extracts the related content identifier of the related content;
The second content identifier is read from the browsing history storage means, the second content identifier having the same content identifier as the related content identifier is searched, and the browsing popularity indicating the frequency of browsing the content of the related content identifier is obtained. Browsing history search means to calculate,
Content processing means for processing the content based on the browsing popularity;
Content output means for outputting the processed content;
An information retrieval apparatus comprising: The browsing history search means calculates the browsing popularity and extracts a related content identifier of related content outside the content indirectly related to the content,
The information search apparatus according to claim 16, wherein the content output means further outputs a related content identifier outside the content. The browsing history storage unit further stores a transition source content identifier of a transition source content that was browsed before the content of the second content identifier stored in the browsing history storage unit was browsed. ,
The browsing history search means includes:
The second content identifier is read from the browsing history storage means, a content identifier having the same content identifier as the received first content identifier is searched, the transition source content identifier corresponding to the content identifier and the transition source content identifier Create a transition source content identifier list that correlates with the number of
A transition source content identifier having the same transition source content identifier as the transition source content identifier in the transition source content identifier list is searched from the browsing history storage means, and the searched transition source content identifier and the number of the transition source content identifiers are obtained. Create an associated transition destination content identifier list,
Using the transition source content identifier list and the transition destination content identifier list, the importance of the transition destination content identifier is calculated,
The information search apparatus according to claim 17, wherein the transition destination content identifier is extracted as a related content identifier outside the content. The browsing history storage unit further stores a transition source content identifier of a transition source content that was browsed before the content of the second content identifier stored in the browsing history storage unit was browsed. ,
The browsing history search means includes:
The second content identifier is read from the browsing history storage means, a content identifier having the same content identifier as the received first content identifier is searched, the transition source content identifier corresponding to the content identifier and the transition source content identifier Create a list that associates the number of
Delete the related content identifier in the content from the transition destination content identifier list,
18. The information search according to claim 17, wherein the transition destination content identifier in the transition destination content identifier list after deletion is extracted as a related content identifier outside the content, and the number of the transition destination content identifiers is set as the importance. apparatus.   The information search device according to claim 18 or 19, wherein the content output means outputs a related content identifier outside the content according to the importance and / or together with the importance. Associating the content with a first content identifier that can identify the content and storing the content in the content storage means;
Storing the second content identifier of the browsed content in the browsing history storage means;
Receiving a search term to search for content;
The content and the first content identifier are read from the content storage means, the content including the search word and the first content identifier corresponding to the content are searched, and the frequency at which the search word appears in the content is determined. Calculating the sentence suitability shown;
The second content identifier is read from the browsing history storage means, a second content identifier having the same content identifier as the searched first content identifier is searched, and the frequency at which the content of the first content identifier is browsed is determined. Calculating the browsing popularity shown,
Synthesizing the sentence fitness and the browsing popularity;
Outputting the retrieved first content identifier based on the synthesized result;
A method for retrieving information, comprising: Storing the second content identifier of the browsed content in the browsing history storage means;
Receiving a first content identifier for desired content;
Obtaining the content of the first content identifier;
Analyzing the related content in the content directly related to the content from the acquired content and extracting the related content identifier of the related content;
The second content identifier is read from the browsing history storage means, the second content identifier having the same content identifier as the related content identifier is searched, and the browsing popularity indicating the frequency of browsing the content of the related content identifier is obtained. A calculating step;
Processing the content based on the browsing popularity;
Outputting the processed content;
A method for retrieving information, comprising: A process of associating the content with a first content identifier that can identify the content and storing the content in the content storage unit;
Processing for storing the second content identifier of the browsed content in the browsing history storage means;
Receiving a search term to search for content;
The content and the first content identifier are read from the content storage means, the content including the search word and the first content identifier corresponding to the content are searched, and the frequency at which the search word appears in the content is determined. A process of calculating the sentence suitability shown,
The second content identifier is read from the browsing history storage means, a second content identifier having the same content identifier as the searched first content identifier is searched, and the frequency at which the content of the first content identifier is browsed is determined. Processing to calculate the viewing popularity shown,
A process of combining the sentence suitability and the browsing popularity;
A process of outputting the searched first content identifier based on the synthesized result;
An information search program for causing a computer to execute the above. Processing for storing the second content identifier of the browsed content in the browsing history storage means;
Receiving a first content identifier for the desired content;
Processing for acquiring the content of the first content identifier;
Analyzing the related content in the content directly related to the content from the acquired content, and extracting the related content identifier of the related content;
The second content identifier is read from the browsing history storage means, the second content identifier having the same content identifier as the related content identifier is searched, and the browsing popularity indicating the frequency of browsing the content of the related content identifier is obtained. Processing to calculate,
Processing the content based on the browsing popularity;
A process of outputting the processed content;
An information search program for causing a computer to execute the above. A process of associating the content with a first content identifier that can identify the content and storing the content in the content storage unit;
Processing for storing the second content identifier of the browsed content in the browsing history storage means;
Receiving a search term to search for content;
The content and the first content identifier are read from the content storage means, the content including the search word and the first content identifier corresponding to the content are searched, and the frequency at which the search word appears in the content is determined. A process of calculating the sentence suitability shown,
The second content identifier is read from the browsing history storage means, a second content identifier having the same content identifier as the searched first content identifier is searched, and the frequency at which the content of the first content identifier is browsed is determined. Processing to calculate the viewing popularity shown,
A process of combining the sentence suitability and the browsing popularity;
A process of outputting the searched first content identifier based on the synthesized result;
A computer-readable recording medium on which is recorded an information search program characterized in that the computer is executed. Processing for storing the second content identifier of the browsed content in the browsing history storage means;
Receiving a first content identifier for the desired content;
Processing for acquiring the content of the first content identifier;
Analyzing the related content in the content directly related to the content from the acquired content, and extracting the related content identifier of the related content;
The second content identifier is read from the browsing history storage means, the second content identifier having the same content identifier as the related content identifier is searched, and the browsing popularity indicating the frequency of browsing the content of the related content identifier is obtained. Processing to calculate,
Processing the content based on the browsing popularity;
A process of outputting the processed content;
A computer-readable recording medium on which is recorded an information search program characterized in that the computer is executed.

Images