JP4498342B2 - Information recommendation system and information recommendation program - Google Patents

Description

  The present invention relates to an information recommendation system and an information recommendation program for recommending an article as a favorite article candidate required by a user to the user.

  In recent years, as one of methods for managing information, there is known a method for registering and managing favorite information so that information required by a user (hereinafter referred to as favorite information) can be accessed instantaneously. In this case, it is managed by holding information for accessing favorite information (hereinafter referred to as access information).

  For example, a system for searching for information useful to a user using a record of a person's value judgment such as reputation is disclosed (see, for example, Patent Document 1).

  In general, in a function called a favorite of a WEB browser, the homepage (favorite information) can be viewed immediately without searching again from the homepage that the user has found (obtained) by searching or the like. For example, the homepage can be registered and managed.

For example, the user can newly add favorite information by actively registering information newly acquired by means such as search as favorite information.
JP 2003-108587 A

  However, in the method of actively adding favorite information as described above, the registered favorite information does not change dynamically unless the user registers the favorite information, for example. Moreover, favorite information can be selected only from information acquired by user operation (information that is candidates for favorite information). Moreover, the favorite information acquired by the user is not necessarily the best quality favorite information desired by the user.

  For this reason, even when there is information newer than the already registered favorite information, the latest information cannot be provided. In addition, information that has not been acquired by the user cannot be provided to the user as the target of favorite information. Also, the user does not know the quality difference between favorite information of the same type.

  An object of the present invention is to provide an information recommendation system and an information recommendation program that can recommend an article as a favorite article candidate for a user to the user.

According to one aspect of the present invention, an article storage unit that stores a plurality of articles including character strings created by a plurality of users, and a favorite article that each of the plurality of users needs corresponds to each user. Favorite storage means for storing information, knowledge link information storage means for storing knowledge link information indicating the relationship between a plurality of articles stored in the article storage means, and a plurality of information stored in the article storage means Storage processing means for storing an article designated by the user as a favorite article in the favorite storage means in association with the user, and an appearance frequency of a character string included in the favorite article designated by the user Based on similar search means for searching similar articles similar to the favorite article from the article storage means, Based on the favorite article stored in the favorite storage means or the knowledge link information stored in the knowledge link information storage means, the user's favorite article candidate from among the similar articles searched by the similar search means A candidate selecting means for selecting an article and a recommendation means for recommending an article as a favorite article candidate selected by the candidate selecting means to the user, and the relationship between the plurality of articles is A new article is created by referring to an article stored in the article storage means, and the candidate selection means is a favorite article specified by the user and a similarity searched by the similarity search means For each article, the favorite article or the similar article is the favorite article of the plurality of users. Registration status information indicating the first number stored in the favorite storage means, or the favorite article specified by the user and the similar article searched by the similar search means, refer to the favorite article or the similar article The knowledge link information indicating that a new article has been created is acquired by the acquisition means for acquiring the reference status information indicating the second number stored in the knowledge link information storage means, and acquired by the acquisition means Based on the first number indicated by the registration status information or the second number indicated by the reference status information, the degree to which the favorite article designated by the user is referred to by the plurality of users Calculating a first reference degree indicating a degree of the first or second number, and each of the similar articles Calculating means for calculating a second reference degree indicating the degree of the first or second number of the similar article as a degree referred to by the plurality of users, and the similar article searched by the similarity searching means An information recommendation system is provided that includes candidate determination means for determining a similar article having a second reference degree higher than the first reference degree as an article that is a candidate for the favorite article .

  According to the present invention, an article that is a candidate for a favorite article for a user can be recommended to the user.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

[First Embodiment]
FIG. 1 is a block diagram showing a hardware configuration of a client-server system including an information recommendation system according to the first embodiment of the present invention.

  The client-server system of FIG. 1 is mainly composed of a computer (server computer) 10 and a plurality of client terminals. The plurality of client terminals include a client terminal 20. On the client terminal 20, client software that uses the computer 10 operates. The client software is a browser, for example. A plurality of client terminals including the client terminal 20 are connected to the computer 10 via a network 30 such as a local area network (LAN). In FIG. 1, client terminals other than the client terminal 20 are omitted.

  The computer 10 is connected to an external storage device 40 such as a hard disk drive. The external storage device 40 stores a program 41 executed by the computer 10. The computer 10 and the external storage device 40 constitute an information sharing system 50.

  FIG. 2 is a block diagram mainly showing a functional configuration of the information recommendation system 50 shown in FIG. The information recommendation system 50 includes an article acquisition unit 51, an information recommendation unit 52, and a similarity search unit 53. In the present embodiment, these units 51 to 53 are realized by the computer 10 shown in FIG. 1 executing the program 41 stored in the external storage device 40. This program 41 can be stored in advance in a computer-readable storage medium and distributed. Further, this program 41 may be downloaded to the computer 10 via the network 30.

  The information recommendation system 50 includes an article database (DB) 42, a favorite database (DB) 43, and a knowledge link (KL) information database (DB) 44. In the present embodiment, these are stored in the external storage device 40.

  The article acquisition unit 51 acquires, for example, article data (hereinafter simply referred to as an article) indicating an article searched according to a search condition specified by the user. Here, the article database 42 stores a plurality of articles including character strings created by a plurality of users, for example. Each of the plurality of articles stored in the article database 42 includes information indicating the date and time when the article was created or updated.

  The information recommendation unit 52 includes an interface 521, a favorite registration unit 524, and a favorite recommendation unit 525.

  The interface 521 inputs and outputs data between client terminals including the client terminal 20 and the information recommendation system 50. The interface 521 includes a favorite registration interface 522 and a favorite candidate output interface 523.

  The favorite registration interface 522 inputs, for example, articles specified by the user operating the client terminal 20 among the articles acquired by the article acquisition unit 51. The favorite candidate output interface 523 outputs to the client terminal 20 articles that are candidates for articles (hereinafter referred to as favorite articles) required by the user operating the client terminal 20 recommended by the favorite recommendation unit 525.

  The favorite registration unit 524 registers (stores) the article input by the favorite registration interface 522 in the favorite database 43 as the favorite article of the user who specified the article. The favorite registration unit 524 registers a favorite article by storing, for example, access information for accessing a favorite article in the favorite database 43. In the favorite article database 43, a favorite article (access information for accessing) is stored in association with each user who designates the favorite article.

  The favorite recommendation unit 525, for example, of articles similar to an article (favorite article) specified by a user operating the client terminal 20 (hereinafter referred to as a similar article) is an article that is a candidate for a favorite article to the user. To recommend. The favorite recommendation unit 525 executes a recommendation process with reference to the favorite database 43 and the KL information database 44. The KL information database 44 stores knowledge link information (hereinafter referred to as KL information) indicating the relationship between a plurality of articles stored in the article database 42. The relationship between a plurality of articles includes, for example, that a new article (reference destination article) is created with reference to an article (reference source article) stored in the article database 42.

  The similarity search unit 53 searches the article database 42 for similar articles of a favorite article specified by, for example, a user operating the client terminal 20. At this time, the similarity search unit 53 executes a search process according to, for example, the appearance frequency of the character string included in the favorite article.

  FIG. 3 is a block diagram showing a configuration of the favorite recommendation unit 525 shown in FIG. The favorite recommendation unit 525 includes a similarity search processing unit 526, a DB processing unit 527, and a favorite candidate selection unit 528.

  The similarity search processing unit 526 executes processing for causing the similarity search unit 53 to search for a similar article of a favorite article specified by the user. The similarity search processing unit 526 receives the similar articles searched by the similarity search unit 53.

  For each similar article received by the favorite article specified by the user and the similar search processing unit 526, the DB processing unit 527 determines that each of the favorite article or the similar article is a user other than the user (hereinafter referred to as another user). Information indicating the number (first number) stored in the favorite database 43 as a favorite article (notation) is acquired. This information indicates a situation in which each of the similar articles received by the favorite article and similar search processing unit 526 is registered as a favorite article of another user. Hereinafter, this information is referred to as registration status information.

  In addition, the DB processing unit 527 refers to each of the favorite articles or similar articles received by the favorite article specified by the user and the similar articles received by the similar search processing unit 526, and a new article has been created. KL information indicating the number (second number) stored in the KL information database 44 is acquired. That is, the DB processing unit 527 acquires information indicating the number of KL information in which each favorite article or similar article corresponds to a reference source article. This information indicates a situation in which each of the favorite articles specified by the user and the similar articles received by the similar search processing unit 526 is referred to by other users. Hereinafter, this information is referred to as reference status information.

  The favorite candidate selection unit 528 selects a favorite article from the similar articles received by the similar search processing unit 526 based on the favorite articles registered in the favorite database 43 or the KL information stored in the KL information database 44. Select candidate articles. An article selected by the favorite candidate selection unit 528 is recommended to the user.

  The favorite candidate selection unit 528 includes a reference degree calculation unit 529 and a candidate determination unit 530. The reference degree calculation unit 529 calculates a first reference degree indicating the degree to which the favorite article is referred to by other users based on the registration state information or the reference state information acquired by the DB processing unit 527. Further, the reference degree calculation unit 529 calculates a second reference degree indicating the degree to which each similar article is referred to by other users based on the registration situation information or the reference situation information acquired by the DB processing unit 527. calculate.

  The candidate determination unit 530 determines an article that is a candidate for a favorite article of the user who operates the client terminal 20 based on the first and second reference degrees calculated by the reference degree calculation unit 529. The candidate determination unit 530 compares the first reference degree and the second reference degree calculated by the reference degree calculation unit 529. The candidate determination unit 530 determines, among the similar articles received by the similarity search processing unit 526, similar articles with a second reference degree higher than the first reference degree as articles that are candidates for the user's favorite articles. The candidate determination unit 530 recommends (notifies) the determined article to a user who operates the client terminal 20, for example, via the candidate output interface 523.

  FIG. 4 shows an example of the data structure of the favorite database 43. In the favorite database 43, a favorite article ID for uniquely identifying a favorite article is stored in association with each user (indicating user ID) designated as the favorite article. Although omitted in FIG. 4, the favorite database 43 stores, for example, for each favorite article, the date (information indicating) that the favorite article was registered.

  In the example illustrated in FIG. 4, the favorite article ID “1” is stored in association with the user ID “A”. Also, a favorite article ID “2” is stored in association with the user ID “B”. Each user can register a plurality of favorite articles in the favorite database 43.

  FIG. 5 shows an example of the data structure of the KL information database 44. The KL information database 44 includes, for each KL information, an original article ID that identifies an article that is a source of a relationship between articles, a previous article ID that identifies an article that is a destination of the relationship between the articles, and the original article. The relationship between the ID and the previous article ID is shown.

  In the example illustrated in FIG. 5, the KL information 441 indicates the original article ID “1”, the previous article ID “2”, and “reference” as the relationship. The KL information 442 indicates the original article ID “3”, the previous article ID “4”, and “reference” as the relationship. When the relationship between the articles is “reference”, the articles indicated by the original article IDs “1” and “3” are the above-described reference source articles. Further, the articles indicated by the previous article IDs “2” and “4” are the aforementioned referenced articles.

  In the example illustrated in FIG. 5, the case of “reference” is described as the relationship between articles. However, for example, “reply”, “revision”, “summary”, and the like may be included in addition to “reference”. .

  Next, the processing procedure of the favorite recommendation unit 525 will be described with reference to the flowchart of FIG.

First, the similarity search processing unit 526 of the favorite recommendation unit 525 acquires a favorite article designated by the user (step S1). This favorite article is an article registered in the favorite database 43 by the favorite registration unit 524. ***
Next, the similarity search processing unit 526 causes the similarity search unit 53 to search for similar articles of the acquired favorite articles (step S2). At this time, the similarity search unit 53 searches for a similar article from the article database 42. The similarity search processing unit 526 acquires the search result searched by the similarity search unit 53.

  The similarity search processing unit 526 determines whether there is a similar article of a favorite article based on the search result by the similarity search unit 53 (step S3).

  When the similar search processing unit 526 determines that a similar article exists (YES in step S3), the DB processing unit 527 searches the favorite article acquired by the similar search processing unit 526 and the similar article searched by the similar search unit 53. Each of the registration status information and the reference status information is acquired (step S4). The DB processing unit 527 refers to the favorite database 43 and the KL information database 44 and acquires registration status information and reference status information.

  The reference degree calculation unit 529 includes the number of favorite articles and similar articles indicated by the registration status information acquired by the DB processing unit 527 registered in the favorite database 43 as favorite articles, or the favorite articles indicated by the reference status information or The degree of reference is calculated based on the number of pieces of KL information in which the similar article corresponds to the reference source article stored in the KL information database 44 (step S5). The reference degree indicates the degree to which an article is referenced by another user.

  Here, the reference degree calculation method will be described in detail. Note that an article (favorite article or similar article) for which a reference degree is calculated will be described as an applicable article. For example, as an example of a calculation formula for obtaining a reference degree, reference degree = {(number of registered articles as favorite articles / number of registered all favorite articles) + (corresponding article corresponds to reference source article) Number of KL information / total number of KL information)} / 2. The total number of registered favorite articles refers to the favorite articles registered in the favorite database 43. The total number of KL information refers to the number of KL information stored in the KL information database 44. Thereby, the reference degree which shows the degree by which the applicable article is referred by the other user can be calculated | required.

  Next, the candidate determination unit 530 compares the reference degree (first reference degree) of the favorite article and the reference degree (second reference degree) of each similar article obtained by the above-described formula, for example. Thereby, the candidate determination unit 530 determines whether or not there is a similar article having a reference degree higher than that of the favorite article (step S6).

  If it is determined that there is a similar article with a higher reference level than the reference level of the favorite article (YES in step S6), the similar article is recommended as a favorite article candidate to the user via the favorite candidate output interface 523 ( Step S7).

  On the other hand, if it is determined in step S3 that the similar search processing unit 526 determines that there is no similar article, the process is terminated, and an article that is a favorite article candidate is not recommended to the user.

  If it is determined in step S6 that there is no similar article having a higher reference level than the reference level of the favorite article, the process is terminated and an article that is a favorite article candidate is not recommended to the user.

  Next, with reference to FIG. 7, a flow of processing for recommending an article as a favorite article candidate to, for example, the user A will be described in detail.

  First, the article acquisition unit 51 acquires an article searched according to, for example, a search condition specified by the user (step S101). In the example illustrated in FIG. 7, the article acquisition unit 51 searches for an article including the keyword “proposal”, for example, and acquires the searched article.

  Here, it is assumed that the article 101 is designated by the user A among the articles acquired by the article acquisition unit 51 and registered as a favorite article (step S102).

  In this case, the similarity search unit 53 searches for similar articles of the favorite article 101. The reference degree calculation unit 529 calculates the reference degree of the similar articles searched by the favorite article 101 and the similarity search unit 53 (step S103). In the example illustrated in FIG. 7, the reference degree of the favorite article 101 is “0.211”. Here, similar articles having a higher reference degree than the favorite article 101 are the article 102 and the article 103.

  Therefore, the article 102 and the article 103 are recommended to the user A as a favorite article candidate. Accordingly, if there is an article desired by the user A among the recommended articles 102 and 103, the user A can register the article (for example, the article 102) as a favorite article (step S104).

  Next, the processing procedure of the favorite recommendation unit 525 will be specifically described with reference to FIG.

  First, the similar search processing unit 526 searches for similar articles of favorite articles (step S201). Hereinafter, the article A among the similar articles searched for will be described.

  Next, the DB processing unit 527 refers to the favorite database 43 and acquires registration status information indicating the number of articles A registered as favorite articles of other users (step S202). In the example illustrated in FIG. 8, the number of articles A registered as favorite articles of other users is four.

  The DB processing unit 527 refers to the KL information database 44 and acquires reference status information indicating the number of KL information in which the article A corresponds to the reference source article (step S203). In the example illustrated in FIG. 8, the number of KL information in which the article A corresponds to the reference source article is 11.

  Candidate determination unit 530 sets the number of articles A registered as a favorite article of other users (registration status information indicating) and the number of KL information corresponding to article A as a reference source article (reference status information indicating). Based on this, the reference degree of the article A is calculated (step S204). Here, it is assumed that the reference degree of the article A is calculated using the above-described formula for obtaining the reference degree. For example, the total number of favorite articles registered is 2000, the number of articles A registered as favorite articles is 1784, the total number of KL information is 1000, and the number of KL information corresponding to article A as a reference source article. Is 466. In this case, the reference degree of the article A is 0.679.

  The candidate determination unit 530 determines whether or not the article A is a favorite article candidate by comparing the reference degree of the favorite article and the reference degree of the article A (step S205). For example, when the reference degree (0.679) of the article A is higher than the reference degree of the favorite article, the candidate determination unit 530 recommends the article A as a favorite article candidate to the user.

  As described above, in the present embodiment, the favorite recommendation unit 525 refers to the favorite database 43 and the KL information database 44, calculates the reference degree of the favorite article and similar articles of the favorite article, and based on the reference degree. To determine articles that are candidates for favorite articles. As a result, articles that are candidates for favorite articles for the user can be recommended to the user. Since the user can register as a favorite article from the recommended articles, for example, even an article other than an article acquired by the user searching the article by operating the client terminal 20 is registered as a favorite article. It becomes possible to do.

[Second Embodiment]
Next, a second embodiment of the present invention will be described with reference to FIG. Note that the hardware configuration of the client-server system including the information recommendation system according to the present embodiment is the same as that of the first embodiment described above, and a description thereof will be omitted.

  FIG. 9 is a block diagram mainly showing a functional configuration of the information recommendation system 60 according to the present embodiment. The same parts as those in FIG. 2 are denoted by the same reference numerals, and detailed description thereof is omitted. Here, parts different from FIG. 2 will be mainly described.

  In the information recommendation system 60 according to the present embodiment, the information recommendation unit 61 replaces the interface 521, the favorite registration unit 524, and the favorite recommendation unit 525 that are included in the information recommendation unit 52 of the first embodiment described above with the interface 611, A login processing unit 613 and a favorite recommendation unit 614 are included.

  The interface 611 includes a login interface 612. For example, when a user logs in from the client terminal 20, the login interface 612 inputs information on the user (hereinafter referred to as user information). This user information is, for example, a user ID for uniquely identifying the logged-in user.

  The login processing unit 613 selects, for example, a favorite article registered as a favorite article from among the favorite articles registered in the favorite database 43 by the user indicated by the user information input through the login interface 612, and then has a certain period of time. A command (command) is issued to the favorite recommendation unit 614 so as to check whether or not (hereinafter referred to as an update target article) exists.

  The favorite recommendation unit 614 determines whether there is an update target article based on a command from the login processing unit 613. Also, the favorite recommendation unit 614 recommends an article that is a candidate for a favorite article among similar articles of the update target article to a logged-in user. The favorite recommendation unit 525 executes a recommendation process with reference to the favorite database 43 and the KL information database 44.

  FIG. 10 is a block diagram illustrating a configuration of the favorite recommendation unit 614. The same parts as those in FIG. 3 are denoted by the same reference numerals, and detailed description thereof is omitted. Here, parts different from FIG. 3 will be mainly described.

  The favorite recommendation unit 614 includes a recommendation execution determination unit 615 and a similar article specification unit 616 in addition to the configuration of the favorite recommendation unit 525 of the first embodiment described above.

  The recommendation execution determination unit 615 determines whether there is a favorite article registered by the logged-in user in the favorite database 43 in accordance with an instruction from the login processing unit 613. When it is determined that there is a favorite article registered in the favorite database 43, the recommendation execution determining unit 615 determines that based on information indicating the date and time when the favorite article stored in the favorite database 43 is registered. It is determined whether the favorite article registered in the favorite database 43 is an update target article.

  The similar article specifying unit 616 generates a similar article created or updated after the date and time when the update target article determined by the recommendation execution determination unit 615 is created or updated among the similar articles received by the similarity search processing unit 526. Identify. The similar article specifying unit 616 executes processing based on information indicating the date and time when the article included in each of the update target article and the similar article is created or updated.

  Next, the processing procedure of the favorite recommendation unit 614 will be described with reference to the flowchart of FIG.

  First, the recommendation execution determination unit 615 of the favorite recommendation unit 614 acquires user information (user ID) indicating the user when the user logs in, for example (step S11).

  Next, the recommendation execution determination unit 615 causes the DB processing unit 527 to search for a favorite article registered in the favorite database 43 by the user based on the acquired user information (step S12).

  The recommendation execution determination unit 615 determines whether there is a favorite article registered in the favorite database 43 by the user indicated by the acquired user information based on the search result by the DB processing unit 527 (step S13). ).

  When it is determined that there is a favorite article (YES in step S13), the recommendation execution determination unit 615 determines whether each of the favorite articles is an update target article (step S14).

  When it is determined that the favorite article is an update target article (YES in step S14), the similarity search processing unit 526 causes the similarity search unit 53 to search for a similar article for each update target article (step S15).

  Next, the process of step S16 corresponding to step S3 of FIG. 6 is executed.

  When the similar search processing unit 526 determines that there is a similar article (YES in step S16), the similar article specifying unit 616 has a date when the update target article is created or updated later. An article is specified (step S17). At this time, the similar article specifying unit 616 executes processing based on information indicating the date and time of creation or update included in each of the update target article and the similar article.

  The DB processing unit 527 performs the process of step S18 corresponding to step S4 of FIG. 6 on the similar article specified by the update target article and the similar article specifying unit 616.

  Next, steps S19 to S21 corresponding to steps S5 to S7 in FIG. 6 are executed.

  On the other hand, if it is determined in step S13 that there is no favorite article, the process is terminated, and an article that is a candidate for a favorite article is not recommended to the user.

  If it is determined in step S14 that the favorite article is not an update target article, the process ends, and an article that is a candidate for a favorite article is not recommended to the user.

  As described above, in the present embodiment, the favorite recommendation unit 614 is created or updated from the update target article among similar articles similar to the update target article registered in the favorite database 43 and for which a certain period of time has passed. The reference degree of similar articles whose date and time is later is calculated, and articles that are candidates for favorite articles are determined based on the reference degrees. As a result, an article that is newer and higher in quality than an update target article that has been registered for a certain period of time can be recommended to the user as a favorite article candidate.

  Note that the present invention is not limited to the above-described embodiments as they are, and can be embodied by modifying the constituent elements without departing from the scope of the invention in the implementation stage. Further, various inventions can be formed by appropriately combining a plurality of constituent elements disclosed in the above embodiments. For example, some components may be deleted from all the components shown in the embodiment. Furthermore, you may combine the component covering different embodiment suitably.

The block diagram which shows the hardware constitutions of the client-server system containing the information recommendation system which concerns on the 1st Embodiment of this invention. The block diagram which mainly shows the function structure of the information recommendation system 50 shown in FIG. The block diagram which shows the structure of the favorite recommendation part 525 shown in FIG. The figure which shows an example of the data structure of the favorite database. The figure which shows an example of the data structure of KL information database. The flowchart which shows the process sequence of the favorite recommendation part 525. The figure for demonstrating concretely the flow of the process which recommends the article used as the candidate of a favorite article. The figure for demonstrating concretely the process sequence of the favorite recommendation part 525. FIG. The block diagram which mainly shows a function structure of the information recommendation system 60 which concerns on the 2nd Embodiment of this invention. The block diagram which shows the structure of the favorite recommendation part 614 shown in FIG. The flowchart which shows the process sequence of the favorite recommendation part 614.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 10 ... Computer, 20 ... Client terminal, 30 ... Network, 40 ... External storage device, 42 ... Article database, 43 ... Favorite database, 44 ... KL information database, 50 ... Information recommendation system, 51 ... Article acquisition part, 52 ... Information Recommendation part 53 ... Similarity search part 60 ... Information recommendation system 61 ... Information recommendation part 521 ... Interface 522 ... Favorite registration interface 523 ... Favorite candidate output interface 524 ... Favorite registration part 525 ... Favorite recommendation part, 526 ... Similarity search processing unit, 527 ... DB processing unit, 528 ... Favorite candidate selection unit, 529 ... Reference degree calculation unit, 530 ... Candidate determination unit, 611 ... Interface, 612 ... Login interface, 613 ... Login processing unit, 614 ... Favorite recommendation section, 6 5 ... recommendation execution determination unit, 616 ... similar articles specifying unit.

Claims (5)

Article storage means for storing a plurality of articles including character strings created by a plurality of users;
Favorite storage means for storing favorite articles required by each of the plurality of users in association with each user,
Knowledge link information storage means for storing knowledge link information indicating a relationship between a plurality of articles stored in the article storage means;
Storage processing means for storing an article designated by a user as a favorite article among the plurality of articles stored in the article storage means in association with the user and storing the article in the favorite storage means;
Similarity search means for searching similar articles similar to the favorite article from the article storage means based on the appearance frequency of the character string included in the favorite article specified by the user;
Based on the favorite article stored in the favorite storage means or the knowledge link information stored in the knowledge link information storage means, the user's favorite article candidate from the similar articles searched by the similarity search means Candidate selection means for selecting articles to become,
A recommendation means for recommending an article as a favorite article candidate selected by the candidate selection means to the user ;
The relationship between the plurality of articles includes that a new article has been created with reference to the articles stored in the article storage means,
The candidate selecting means is
For each favorite article specified by the user and similar articles searched by the similarity search means, the favorite article or the similar articles are stored in the favorite storage means as favorite articles of the plurality of users. The registration status information indicating the number or the favorite article specified by the user and the similar article searched by the similar search means that a new article has been created with reference to the favorite article or the similar article. Acquisition means for acquiring reference status information indicating the second number stored in the knowledge link information storage means.
Based on the first number indicated by the registration status information acquired by the acquisition means or the second number indicated by the reference status information, the favorite article designated by the user is referenced by the plurality of users. A first reference degree indicating the degree of the first or second number for the favorite article as a degree is calculated, and the degree for the similar article is calculated as a degree that each of the similar articles is referred to by the plurality of users. Calculating means for calculating a second reference degree indicating the degree of the first or second number;
Candidate determination means for determining a similar article having a second reference degree higher than the first reference degree among the similar articles searched by the similarity search means as an article that is a candidate for the favorite article;
including
An information recommendation system characterized by this. An input means for inputting user information indicating the user according to a user operation;
A recommendation execution determination unit that determines whether there is a favorite article stored in the favorite storage unit in association with the user indicated by the user information input by the input unit;
The information recommendation system according to claim 1, wherein the similarity search unit searches for a similar article similar to the favorite article when the recommendation execution determination unit determines that a favorite article exists. The storage processing means includes information indicating the date and time when a favorite article specified by a user among a plurality of articles stored in the article storage means is associated with the user and stored in the favorite storage means. Store in storage means,
When it is determined that the favorite article exists, the recommendation execution determination unit stores the favorite article in the favorite information storage unit based on information indicating a date and time when the favorite article is stored in the favorite storage unit. Update determination means for determining whether a certain period of time has passed since
The information recommendation system according to claim 2, wherein the similarity search unit searches for a similar article similar to a favorite article determined by the update determination unit that a certain period of time has passed. Each of the plurality of articles stored in the article storage means includes information indicating the date and time when the article was created or updated,
The candidate selecting means is
Based on the information indicating the date and time when the article included in each of the favorite articles and the similar articles searched by the similar search means determined that the fixed period has passed by the update determination means, A specifying unit for specifying a similar article created or updated after the date and time when the favorite article was created or updated among the similar articles searched by the similar search unit;
Based on the favorite article stored in the favorite storage means or the knowledge link information stored in the knowledge link information storage means, the article as a candidate of the favorite article among the similar articles specified by the specifying means The information recommendation system according to claim 3, wherein the information recommendation system is selected. Article storage means for storing a plurality of articles including character strings created by a plurality of users, favorite storage means for storing favorite articles required by each of the plurality of users in association with each user, and Information recommendation comprising an external storage device having knowledge link information storage means for storing knowledge link information indicating the relationship between a plurality of articles stored in the article storage means, and a computer using the external storage device An information recommendation program executed by the computer in the system,
In the computer,
Of the plurality of articles stored in the article storage means, an article specified by a user is stored in the favorite storage means in association with the user as a favorite article;
Searching similar articles similar to the favorite article from the article storage unit based on the appearance frequency of the character string included in the favorite article specified by the user;
Based on the favorite articles stored in the favorite storage means or the knowledge link information stored in the knowledge link information storage means, articles that are candidates for the user's favorite articles from the searched similar articles. A determining step;
Recommending the article as a candidate for the determined favorite article to the user ,
The relationship between the plurality of articles includes that a new article has been created with reference to the articles stored in the article storage means,
The step of determining an article that is a candidate for the user's favorite article includes:
For each favorite article specified by the user and the searched similar articles, a registration indicating the first number of the favorite articles or the similar articles stored in the favorite storage means as the favorite articles of the plurality of users. Knowledge link information indicating that a new article has been created with reference to the favorite article or the similar article for each favorite article specified by the user and the searched similar article is the knowledge link. Obtaining reference status information indicating the second number stored in the information storage means;
Based on the first number indicated by the acquired registration status information or the second number indicated by the reference status information, the degree to which the favorite article specified by the user is referred to by the plurality of users A first reference degree indicating a degree of the first or second number for the favorite article is calculated, and the first or second for the similar article is calculated as a degree that each of the similar articles is referred to by the plurality of users. Calculating a second reference degree indicating a degree of a number of two;
Determining a similar article with a second reference degree higher than the first reference degree among the retrieved similar articles as an article that is a candidate for the favorite article;
including
An information recommendation program characterized by that.

Images