JP2015130110A - Route search device, program and route search system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To present a route connecting two persons different in a standpoint or environment.SOLUTION: A route search device comprises: a submission data analysis unit 1 that collects a submission to an SNS, extracts a plurality of keywords; selects any one of the keywords, calculates a level of relevancy between the selected specific keyword and a plurality of other keywords, sorts out the plurality of other keywords in descending order of the level of relevancy, and creates relevancy data for recording for each specific keyword; and a route search unit 2 that, when a first and input keywords from a user exist as the keyword of any of the recording relevancy data, collates first collation relevancy data with second collation relevancy data, and when the same keyword exists in the first and second collation relevancy data, outputs a route from the first input keyword via the same keyword to the second input keyword.

Description

  The present invention relates to a technique for extracting a relationship between any two parties based on content posted to an SNS (Social Networking Service) and presenting a route between the two parties.

  Recently, with the spread of SNS, people who have a common hobby on the net but are not acquainted with each other form a community. On the other hand, in the real world, it is not easy for people who are not acquainted with each other to communicate smoothly. For this reason, studies have been conventionally conducted to present common information in order to activate communication between two parties.

  As common information for activating communication, a system has been proposed in which user information of a contributor is extracted from posted content on social media and the common information of the two parties is presented or a potential shared relationship is estimated.

  For example, Non-Patent Document 1 proposes a system that presents common information such as “soccer” from user-related information between two parties in the same space. The user-related information targets nouns such as personal information and hobbies among user information obtained from the social graph.

  Further, Non-Patent Document 2 proposes a social graph expression method that estimates a potential shared relationship of people obtained by a social graph and represents a mutual relationship. In Non-Patent Document 2, when using a social graph, not only the relationship between people but also social events and communities are included as nodes. A potential common relationship is a node other than a person.

Information Processing Society of Japan Interaction 2011 MAKOTO: Proposal of communication support system using social graph IEICE Technical Report NS2010-156 A social graph expression method for estimating potential shared relationships between people

  However, in Non-Patent Document 1, since common information is presented from user-related information between two parties in the same space, in the case of two parties in completely different environments and positions, they are not necessarily common to both. There is a problem that it cannot be said that information exists, and smooth communication between the two parties in the real world cannot be expected.

  In Non-Patent Document 2, adding nodes such as social events, communities, etc. to two people who have a common acquaintance but have no direct connection on the social graph with only the relationship between people as nodes. And they discover and recommend direct common information other than people. That is, Non-Patent Document 2 does not target a relationship between two parties in completely different environments or positions, and it is difficult to present a route between the two parties.

  The present invention has been made in view of such circumstances, and an object of the present invention is to provide a route search device, a program, and a route search system capable of presenting a route connecting two persons having different positions and environments. .

  (1) In order to achieve the above object, the present invention takes the following measures. That is, the route search device of the present invention is a route search device that extracts the relationship between any two parties based on the contents posted to SNS (Social Networking Service) and presents the route between the two parties. The SNS collects posts to the SNS, performs a morphological analysis process on the collected posts, extracts a plurality of keywords, selects any one keyword from the plurality of extracted keywords, and selects the selected specific keyword and A post data analysis unit that calculates a degree of association with another plurality of keywords, rearranges the plurality of other keywords in descending order of degree of association with the specific keyword, and creates recording relation data for each of the specific keywords; The first input keyword and the second input keyword input by the user exist as specific keywords of any one of the recording related data. If the first input keyword exists as a specific keyword of any of the recording related data, the recording related data with the first input keyword as the specific keyword is used for the first verification. When the second input keyword exists as a specific keyword of any of the recording related data, the recording related data having the second input keyword as the specific keyword is stored in the database as related data. Are stored in a database as collation relation data for the first collation, the first collation relation data and the second collation relation data are collated, and the first collation relation data and the second collation relation data. If the same keyword is present in the first input keyword passing through the same keyword, A route searching unit for outputting a path to the input keywords, characterized in that it comprises a.

  In this way, posts to SNS are collected, morpheme analysis processing is performed on the collected posts, a plurality of keywords are extracted, one keyword is selected from the plurality of extracted keywords, the selected specific keyword and others The degree of association with a plurality of keywords is calculated, the other plurality of keywords are sorted in descending order of the degree of association with a specific keyword, and recording relation data is created for each specific keyword. It becomes possible to find common information other than users (people), and based on this common information, it is possible to construct a database for connecting people who are not acquainted with each other. Further, it is searched whether or not the first input keyword and the second input keyword input from the user exist as specific keywords of any recording relation data, and the first input keyword is any recording relation. When it exists as a specific keyword of data, the recording relation data having the first input keyword as the specific keyword is stored in the database as the first collation relation data, and the second input keyword is one of the recording relations. When it exists as a specific keyword of data, the relational data for recording having the second input keyword as the specific keyword is stored in the database as the second collating relational data, and the first collating relational data and the second collating data are stored. The related data is collated, and the same key is included in the first collating relation data and the second collating relation data. If the word is present, since the output path from the first input keyword through the same keywords to the second input keyword, it is possible to create an opportunity to connect the human together with no acquaintance with each other.

  (2) Moreover, the route search device of the present invention selects any one user name as a specific keyword on the SNS, and at least one other user associated with the specific keyword on the SNS. A social graph analysis unit that extracts a name, attaches a certain degree of association to the extracted user name, and generates SNS relation data including other user names with the certain degree of association for each specific keyword; Integrating the SNS relation data into the recording relation data, rearranging other keywords and other user names in descending order of the degree of relation, and generating relation data for each specific keyword And a section.

  In this manner, on the SNS, any one user name is selected as a specific keyword, and at least one other user name related to the specific keyword on the SNS is extracted, and the extracted user name is constant. SNS relation data including other user names with a certain degree of relation attached is generated for each specific keyword, and the SNS relation data is integrated with the relation data for recording. Other keywords and other user names are rearranged to generate recording relation data for each specific keyword. Therefore, even users who are already associated on the SNS have different degrees of association according to the content of the posting. Therefore, different routes can be presented according to the degree of association. This makes it possible to obtain the shortest path between the two parties. As a result, it becomes possible to connect people who are not acquainted with each other.

  (3) Further, in the route search device of the present invention, the route search unit, when the same keyword does not exist in the first matching relationship data and the second matching relationship data, A first highest relevance keyword having the highest degree of relevance is extracted from the first relational relation data, and a second highest relevance degree keyword having the highest degree of relevance is extracted from the second relational data for matching. The first highest relevance keyword and the second highest relevance keyword are searched for as a specific keyword of any one of the recording relation data, and the first highest relevance keyword is any When it exists as a specific keyword of the recording relation data, the recording relation data with the first highest relevance keyword as a specific keyword is used as third matching relation data. In the database, and when the second highest relevance keyword exists as a specific keyword of any of the recording related data, the recording related data using the second highest relevance keyword as the specific keyword Stored in the database as fourth collating relational data, collating the third collating relational data with the fourth collating relational data, the third collating relational data and the fourth collating relational data When the same keyword exists in the relational data, the route of the first input keyword, the first highest relevance keyword, the same keyword, the second highest relevance keyword, and the second input keyword is determined. It is characterized by outputting.

  With this configuration, even if the same keyword cannot be found in the first matching, a keyword corresponding to the degree of association is extracted and the second matching is performed to present the shortest path between the two parties. It becomes possible. As a result, it becomes possible to connect people who are not acquainted with each other.

  (4) Further, in the route search device of the present invention, the post data analysis unit calculates the degree of association based on the time required for posting to the post, the number of postings, the posting frequency, or the content of the posted sentence analysis. It is characterized by calculating.

  In this way, since the degree of association is calculated based on the time required for posting, the number of postings, the posting frequency, or the posted text analysis content, the relationship obtained from the posted content or indirectly to the two parties It is possible to build a relationship through the target node.

  (5) Moreover, the route search apparatus of this invention is further provided with the display part which displays the path | route which the said route search part output so that visual recognition is possible.

  In this way, since the route output by the route search unit is displayed so as to be visible, it is possible to present the shortest route between the two, and to connect people who are not acquainted with each other.

  (6) Moreover, the program of this invention extracts the relationship between arbitrary two parties based on the content contributed to SNS (Social Networking Service), and the route search apparatus which shows the path | route between the said two parties A process for collecting posts to SNS, a process for performing morphological analysis on the collected posts and extracting a plurality of keywords, and selecting any one keyword from the plurality of extracted keywords Processing, calculating the degree of association between the selected specific keyword and the plurality of other keywords, rearranging the other plurality of keywords in descending order of the degree of association with the specific keyword, for each of the specific keywords The process for creating the relation data for recording, and the recording relation in which the first input keyword and the second input keyword input by the user are either A process for searching whether or not the data exists as a specific keyword of data, and when the first input keyword exists as a specific keyword of any of the recording related data, recording using the first input keyword as the specific keyword Processing for storing the related data in the database as the first matching related data, and when the second input keyword exists as a specific keyword of any of the recording related data, the second input keyword is specified A process for storing the relation data for recording as a keyword in the database as second collation relation data, a process for collating the first collation relation data with the second collation relation data, and the first If the same keyword exists in the matching relation data and the second matching relation data, the same And outputting a route from the first input keyword to the second input keyword through a keyword, a series of processing, and characterized by causing a computer to execute.

  In this way, posts to SNS are collected, morpheme analysis processing is performed on the collected posts, a plurality of keywords are extracted, one keyword is selected from the plurality of extracted keywords, the selected specific keyword and others The degree of association with a plurality of keywords is calculated, the other plurality of keywords are sorted in descending order of the degree of association with a specific keyword, and recording relation data is created for each specific keyword. It becomes possible to find common information other than users (people), and based on this common information, it is possible to construct a database for connecting people who are not acquainted with each other. Further, it is searched whether or not the first input keyword and the second input keyword input from the user exist as specific keywords of any recording relation data, and the first input keyword is any recording relation. When it exists as a specific keyword of data, the recording relation data having the first input keyword as the specific keyword is stored in the database as the first collation relation data, and the second input keyword is one of the recording relations. When it exists as a specific keyword of data, the relational data for recording having the second input keyword as the specific keyword is stored in the database as the second collating relational data, and the first collating relational data and the second collating data are stored. The related data is collated, and the same key is included in the first collating relation data and the second collating relation data. If the word is present, since the output path from the first input keyword through the same keywords to the second input keyword, it is possible to create an opportunity to connect the human together with no acquaintance with each other.

  (7) Further, the route search system of the present invention includes a server device and a client device, and extracts a relationship between any two parties based on content posted to an SNS (Social Networking Service). A route search system for presenting a route between two parties, wherein the client device receives an input of a keyword from a user and acquires the input keyword from the server and an input unit that transmits the input keyword to the server device A display unit for displaying information, wherein the server device collects posts to the SNS, performs morpheme analysis processing on the collected posts, extracts a plurality of keywords, and selects any of the extracted keywords Select one keyword, calculate the degree of association between the selected keyword and other keywords, A plurality of other keywords are rearranged in descending order of the degree of relevance, a post data analysis unit that creates recording relation data for each specific keyword, and a first input keyword and a second input keyword input by the user A search is made as to whether or not any one of the recording related data exists as a specific keyword, and when the first input keyword exists as any one of the recording related data, the first input keyword is When the recording related data as the specific keyword is stored in the database as the first collation related data, and the second input keyword exists as a specific keyword of any of the recording related data, the second The relation data for recording that uses the input keyword as a specific keyword is the second relation data for verification. And storing the first collation relation data and the second collation relation data in the database, and the same keyword in the first collation relation data and the second collation relation data. A route search unit that transmits a route from the first input keyword to the second input keyword via the same keyword to the client device, the display unit of the client device Displays the route so as to be visible.

  In this way, posts to SNS are collected, morpheme analysis processing is performed on the collected posts, a plurality of keywords are extracted, one keyword is selected from the plurality of extracted keywords, the selected specific keyword and others The degree of association with a plurality of keywords is calculated, the other plurality of keywords are sorted in descending order of the degree of association with a specific keyword, and recording relation data is created for each specific keyword. It becomes possible to find common information other than users (people), and based on this common information, it is possible to construct a database for connecting people who are not acquainted with each other. Further, it is searched whether or not the first input keyword and the second input keyword input from the user exist as specific keywords of any recording relation data, and the first input keyword is any recording relation. When it exists as a specific keyword of data, the recording relation data having the first input keyword as the specific keyword is stored in the database as the first collation relation data, and the second input keyword is one of the recording relations. When it exists as a specific keyword of data, the relational data for recording having the second input keyword as the specific keyword is stored in the database as the second collating relational data, and the first collating relational data and the second collating data are stored. The related data is collated, and the same key is included in the first collating relation data and the second collating relation data. If the word is present, since the output path from the first input keyword through the same keywords to the second input keyword, it is possible to create an opportunity to connect the human together with no acquaintance with each other.

  According to the present invention, two parties that have not been involved so far can be connected through an indirect target node. As a result, people who are not acquainted with each other can smoothly communicate with each other.

It is a block diagram which shows schematic structure of the route search apparatus which concerns on this embodiment. 3 is a block diagram illustrating a schematic configuration of a posted data analysis unit 1. FIG. 3 is a block diagram illustrating a schematic configuration of a route search unit 2. FIG. It is a flowchart which shows operation | movement of this embodiment. It is a block diagram which shows schematic structure of the route search system which concerns on this embodiment.

  The route search device according to the embodiment of the present invention extracts information representing the connection of the poster from the content posted to the SNS (Social Networking Service), and analyzes the information representing the connection in all the extracted information. Then, a relationship obtained from the posted content or a relationship through the target node that is indirect to the two parties is constructed. In the present embodiment, as an example, information representing the connection of posters in the company is used. This information includes, for example, a person name, an organization name, or a professional name.

  FIG. 1 is a block diagram illustrating a schematic configuration of a route search apparatus according to the present embodiment. The route search apparatus according to the present embodiment includes a posted data analysis unit 1, a route search unit 2, a social graph analysis unit 3, and a relational data integration unit 5.

FIG. 2 is a block diagram illustrating a schematic configuration of the posted data analysis unit 1. The morphological analysis unit 101 collects SNS user posts and extracts keywords from the collected user posts. For example, the content posted by contributor u1 received an inquiry from Department B of Company S who was interested in a press release related to the dialogue system " Product P " with the introduction of Mr. A. K. Individual demonstrations and explanations Was done. ", The morphological analysis unit 101 extracts the underlined part as a keyword.

  The first keyword selection unit 103 selects only the keyword representing the connection of the poster u1 from the keywords extracted by the morpheme analysis unit 101. As a keyword representing the connection of the poster, for example, there are a personal name, an organization name, a professional name, and the like. Specific examples are “personal name: Mr. K”, “organization name: Department A, Company S, Department B”, and “special name: product P, dialogue system”.

  The weight calculation unit 105 calculates the degree of association with the poster for all the keywords selected by the first keyword selection unit 103. For personal names, weighting is performed according to the frequency of comments in past postings or the response time of comments, and for other keywords, weighting is performed according to the number of appearances.

Hereinafter, an example of keyword weight calculation for u1 will be shown.
“Mr. K: When the comment from Mr. u1 to Mr. K is 10 times and the average response time is 1h, the weight of Mr. K: 1/10 (times) × 1 (h) = 1/10 = 0.1”
“Part A: When the number of occurrences of Part A in the posting of u1 is 5, the weight of Part A: 1/5 (times) = 0.2”
Note that the smaller the weight, the higher the degree of association.

  The first relationship generation unit 107 generates and holds the relationship data in the descending order of the degree of association with the poster in the weight calculation unit 105 for all keywords selected by the poster and the first keyword selection unit 103.

  For example, the relation data is generated as follows.

(U1; Mr. K: 0.1, Part A: 0.2, Company S: 0.5, Part B: 1, Product P: 3, Dialogue System: 5),
(U2; ...),
...
(Part C; u2: 0.1, u3: 0.3, Part A: 0.5, g2: 1, ...),
(G2; ...),
...
The SocialGraph analysis unit 3 shown in FIG. 1 analyzes the relationship between users (people) in the SNS, and generates based on the friendship on the SNS, which is referred to as friendship / follower relationship on the SNS. For this reason, the SocialGraph analysis unit 3 owns friendship data for all N users registered in the target SNS. Further, it is assumed that the relation data based on the friendship on the SNS is not weighted to indicate the degree of association, and the degree of association is always a constant value.

  For example, when the user u1 has a friendship with u2, u5,..., UN on the SNS, the relational data in the SocialGraph analysis unit 3 of u1 is expressed as follows.

(U1; u2: 1, u5: 1, ..., uN: 1)
...
(UN; u1: 1, ...)
The relational data integration unit 5 shown in FIG. 1 integrates the relational data generated by the SocialGraph analysis unit 3 and the relational data generated by the post data analysis unit 1.

  For example, an example of integration of u1 is as follows.

(U1; Mr. K: 0.1, A part: 0.2, S company: 0.5, B part: 1, u2: 1, u5: 1, ..., uN: 1, product P: 3, dialogue system : 5)
FIG. 3 is a block diagram illustrating a schematic configuration of the route search unit 2. As shown in FIG. 3, the route search unit 2 includes a first route search unit 25, a second route search unit 27, a DB 29, a matching unit 31, a necessary route extraction unit 33, and a route presentation unit 35. .

  The first route search unit 25 includes a first keyword search unit 37a, a database registration unit 39, a second relationship generation unit 41a, and a second keyword selection unit 43a. The second route search unit 27 includes a second keyword search unit 37b, a database registration unit 39, a third relationship generation unit 41b, and a third keyword selection unit 43b. The components of the first route search unit 25 and the second route search unit 27 have the same function. In addition, the DB 29 is shared by the first route search unit 25 and the second route search unit 27.

  The user inputs a “from word” as a route search source and a “to word” as a search destination. The word that can be input is a word representing the connection of the poster. For example, “from: u1, to: C part” can be used as an input keyword.

  The first keyword search unit 37a searches whether the keyword (from word) input by the user on the application exists in the set of relational data generated and held by the posted data analysis unit 1. If it exists, the process proceeds to the database registration process. If it does not exist, the user input is prompted again.

  Similarly, the second keyword search unit 37b searches whether the keyword (to word) input by the user on the application exists in the set of relational data generated and held by the posted data analysis unit 1. . If it exists, the process proceeds to the database registration process. If it does not exist, the user input is prompted again.

  Each database registration unit 39 refers to the relationship data held in the posted data analysis unit 1 and registers the keyword, which is the relationship data of the input keyword, in the database.

  For example, the following contents are registered in the database.

“First route search unit 25: (u1) Mr. K, part A, company S, ...”
"Second route search unit 27: (C part) u2, u3, A part"
The second relation generation unit 41a generates and holds relation data for the input keyword and relation keyword in ascending order of weight. Similarly, the third relation generation unit 41b generates and holds relation data for the input keyword and relation keyword in ascending order of weight. For example, the Dijkstra method may be used as a method for generating relational data in ascending order of weight.

  For example, the following contents are used as related data.

"First route search unit 25 (u1: Mr. K, Part A, Company S, ...)"
“Second route search unit 27 (C unit: u2, u3, A unit,...)”
The collation unit 31 refers to the databases registered by the first route search unit 25 and the second route search unit 27 to check for the presence of the same keyword. If there is the same keyword, the process proceeds to the necessary route extraction unit 33. For example, it is assumed that “A part” exists when the same keyword exists.

  On the other hand, if there is no identical keyword, the first route search unit 25 executes the process in the first keyword search unit 37a again, and the second route search unit 27 performs the process in the second keyword search unit 37b. Try again. Here, a keyword to be connected next is selected from the data related in each process of the second relationship generation unit 41a and the third relationship generation unit 41b, and the first keyword search unit 37a or the second keyword search is performed. It will return to the part 37b.

For example, the following keywords are selected as keywords to be connected next.
Keyword reselection:
First route search unit 25: “Mr. K”
Second route search unit 27: “u2”

  The necessary route extraction unit 33 uses the data stored in the relationship generation processing in the second relationship generation unit 41a of the first route search unit 25 and the third relationship generation unit 41b of the second route search unit 27 based on the data. Only the relational data from the keyword input to the keyword collated by the collation unit 31 is extracted.

For example, the necessary route extraction unit 33 can extract the following route.
First route search unit 25: “u1 → A part”
Second route search unit 27: “C part → A part”

  The route presentation unit 35 combines the two routes extracted by the necessary route extraction unit 33 with a matching keyword, and presents a route from “from word” to “to word” input by the user.

  For example, the route presentation unit 35 presents the following route.

“Route example: u1 → A part → C part”
Further, as described above, when there is no identical keyword in the first collation, the following route may be presented.

“Route example: u1 → K → u2 → C section”
Note that the present invention can be applied not only within a company but also to other cases such as hobbies. It can be applied to other cases by appropriately preparing a dictionary in which information representing the connection of the poster is registered.

  FIG. 4 is a flowchart showing the operation of the present embodiment. First, SNS posts are collected (step S1). Here, the target documents posted to the SNS are collected. At this time, the file is divided for each writer and each community. Next, SNS posting analysis is performed (step S2). The data collected and divided in step S1 is analyzed, and keywords such as personal names, organization names, and specialized names are extracted. Next, relation data is generated (step S3). In this case, the weighted keywords and keywords extracted in step S2 are created and stored as related data.

  Next, a keyword is input from the user (step S4). Here, the user is prompted to input two keywords at both ends of the route that the user wants to know. Next, a route search is performed (step S5). The route is searched while the relational data generated in step S3 and the keyword input in step S4 are collated. Finally, route presentation is performed (step S6). The route is presented to the user by displaying the search result on the screen.

  Note that the route search apparatus according to the present embodiment can also be configured as a server client system. FIG. 5 is a block diagram showing a schematic configuration of the route search system according to the present embodiment. Post data analysis unit 1, route search unit 2, social graph analysis unit 3, and relational data integration unit 5 are mounted on a server device, and a user inputs a keyword from client device 50. The route presented by the server device is output to the client device 50 and displayed on the screen.

  As described above, according to the present embodiment, it is possible to connect two parties that have not been involved so far via an indirect target node. As a result, people who are not acquainted with each other can smoothly communicate with each other.

DESCRIPTION OF SYMBOLS 1 Post data analysis part 2 Path | route search part 3 Social Graph analysis part 5 Relational data integration part 25 1st path | route search part 27 2nd path | route search part 31 Collation part 33 Necessary path | route extraction part 35 Path | route presentation part 37a 1st keyword Search unit 37b Second keyword search unit 39 Database registration unit 41a Second relationship generation unit 41b Third relationship generation unit 43a First keyword selection unit 43b Second keyword selection unit 50 Client device 101 Morphological analysis unit 103 1 keyword selection unit 105 weight calculation unit 107 first relation generation unit

Claims (7)

A route search device that extracts a relationship between any two parties based on content posted to an SNS (Social Networking Service) and presents a route between the two parties,
Collecting posts to SNS, performing morphological analysis on the collected posts, extracting a plurality of keywords, selecting any one keyword from the plurality of extracted keywords, and selecting the selected specific keyword and other plurality of keywords A post data analysis unit that calculates the degree of association with the keyword, rearranges the plurality of other keywords in descending order of degree of association with the specific keyword, and creates recording relation data for each specific keyword;
It is searched whether or not the first input keyword and the second input keyword input by the user exist as specific keywords of any of the recording related data, and the first input keyword is any of the recording keywords. When there is a specific keyword of the relational data, the recording relational data having the first input keyword as the specific keyword is stored in the database as the first collation relational data, and the second input keyword is any When there is a specific keyword of the recording relation data, the recording relation data having the second input keyword as a specific keyword is stored in the database as second verification relation data, and the first verification relation data And the second relational data for collation, the first relational data for collation and the previous A route search unit for outputting a route from the first input keyword to the second input keyword via the same keyword when the same keyword exists in the second matching relation data; A route search apparatus comprising: On the SNS, any one user name is selected as a specific keyword, and at least one other user name associated with the specific keyword on the SNS is extracted, and the extracted user name is constant. A social graph analysis unit that generates an SNS relation data for each specific keyword including an association degree and another user name with the certain degree of association;
Integrating the SNS relation data into the recording relation data, rearranging other keywords and other user names in descending order of the degree of relation, and generating relation data for each specific keyword The route search apparatus according to claim 1, further comprising a unit.   When the same keyword does not exist in the first matching relation data and the second matching relation data, the route search unit has the highest degree of relevance from the first matching relation data. A high first highest relevance keyword is extracted, and a second highest relevance keyword having the highest relevance is extracted from the second matching relation data, and the first highest relevance keyword and the second relevance keyword are extracted. Searching whether the highest relevance keyword exists as a specific keyword of any of the recording related data, and when the first highest relevance keyword exists as a specific keyword of any of the recording related data, The recording relation data having the first highest relevance keyword as a specific keyword is stored in a database as third matching relation data, and When two highest relevance keywords are present as specific keywords of any of the recording related data, the database is set with the recording related data having the second highest relevance keyword as the specific keyword as fourth matching relation data. And the third collating relation data and the fourth collating relation data are collated, and the same keyword is found in the third collating relation data and the fourth collating relation data. The path of the first input keyword, the first highest relevance keyword, the same keyword, the second highest relevance keyword, and the second input keyword if output exists. The route search device according to claim 1 or 2.   The post data analysis unit calculates the degree of association based on time required for posting with respect to a post, the number of posts, the frequency of posting, or the contents of posted text analysis. 4. The route search device according to any one of 3.   The route search device according to any one of claims 1 to 4, further comprising a display unit that displays the route output by the route search unit so as to be visible. A program for a route search apparatus that extracts the relationship between any two parties based on the content posted to SNS (Social Networking Service) and presents the route between the two parties.
Processing to collect posts to SNS;
A process of performing a morphological analysis process on the collected posts and extracting a plurality of keywords;
A process of selecting any one keyword from the plurality of extracted keywords;
Processing for calculating the degree of association between the selected specific keyword and other keywords;
Rearranging the plurality of other keywords in descending order of the degree of association with the specific keyword, and creating recording relation data for each specific keyword;
A process of searching whether or not the first input keyword and the second input keyword input by the user exist as specific keywords of any of the recording related data;
When the first input keyword exists as a specific keyword of any of the recording relation data, the recording relation data having the first input keyword as the specific keyword is stored in the database as the first matching relation data. Processing to
When the second input keyword exists as a specific keyword of any one of the recording relation data, the recording relation data having the second input keyword as the specific keyword is stored in the database as second collation relation data. Processing to
A process of collating the first collating relational data with the second collating relational data;
When the same keyword exists in the first collation relation data and the second collation relation data, the first input keyword through the same keyword to the second input keyword A program for causing a computer to execute a series of processes of outputting a route. A route search system that includes a server device and a client device, extracts a relationship between any two parties based on content posted to an SNS (Social Networking Service), and presents a route between the two parties. And
The client device is
An input unit that accepts an input of a keyword from a user and transmits the input keyword to the server device;
A display unit for displaying information acquired from the server,
The server device
Collecting posts to SNS, performing morphological analysis on the collected posts, extracting a plurality of keywords, selecting any one keyword from the extracted keywords, and selecting the selected specific keyword and other plurality of keywords A post data analysis unit that calculates the degree of association with the keyword, rearranges the plurality of other keywords in descending order of degree of association with the specific keyword, and creates recording relation data for each specific keyword;
It is searched whether or not the first input keyword and the second input keyword input by the user exist as specific keywords of any of the recording related data, and the first input keyword is any of the recording keywords. When there is a specific keyword of the relational data, the recording relational data having the first input keyword as the specific keyword is stored in the database as the first collation relational data, and the second input keyword is any When there is a specific keyword of the recording relation data, the recording relation data having the second input keyword as a specific keyword is stored in the database as second verification relation data, and the first verification relation data And the second relational data for collation, the first relational data for collation and the previous A route for transmitting a route from the first input keyword to the second input keyword via the same keyword to the client device when the same keyword is present in the second matching relation data A search unit,
The route search system, wherein the display unit of the client device displays the route so as to be visible.

Images