JP2017076376A - Calculation device, calculation method and calculation program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reflect actual popularity in priority.SOLUTION: A calculation device according to the present invention comprises: an acquisition unit to acquire action information which indicates the user action history regarding selection of selection targets, and input information which is input by the user; and a calculation unit to calculate priority in presenting a prescribed selection target on the basis of action information indicating the user action history regarding selection of a prescribed selection target of the action information, and input information including information associated with the prescribed selection target of the input information.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a calculation device, a calculation method, and a calculation program.

  Conventionally, electronic commerce via the Internet, such as Internet auctions and EC (Electronic Commerce) sites, has been performed. As an example of such electronic commerce, in order to increase users' willingness to purchase, the popularity of products and services (hereinafter simply referred to as products) is estimated, and products with a high estimated popularity are given priority. The technique to present is known. For example, the priority based on the search result of the product corresponding to the search query is corrected by the priority based on the result of performing a web search for the product, so that the priority reflecting the search tendency related to the search query is changed. A technique for calculating and displaying products in order based on the calculated priority is known.

JP 2010-61420 A

  However, in the technique described above, the actual popularity may not be reflected in the priority. For example, in the above-described technology, the priority based on the search result of the product corresponding to the search query is corrected by the priority based on the result of performing a web search for the product. Products that are searched or purchased on a daily basis may be preferentially presented over products that are actually popular.

  The present application has been made in view of the above, and an object thereof is to provide a calculation device, a calculation method, and a calculation program capable of reflecting an actual popularity degree in a priority.

  The calculation apparatus according to the present application includes an acquisition unit that acquires behavior information indicating a history of a user's behavior related to selection of a selection target, and input information input by the user, and a predetermined selection target among the behavior information. Priority when presenting a predetermined selection target using behavior information indicating a history of user behavior regarding selection and input information including information related to the predetermined selection target among the input information And a calculating unit for calculating

  According to one aspect of the embodiment, there is an effect that the actual popularity can be reflected in the priority.

FIG. 1 is a diagram illustrating an example of an information distribution apparatus according to the embodiment. FIG. 2 is a diagram illustrating an example of a connection relationship between the information distribution apparatus according to the embodiment and another apparatus. FIG. 3 is a diagram illustrating an example of a functional configuration of the information distribution apparatus according to the embodiment. FIG. 4 is a diagram illustrating an example of information registered in the target information database according to the embodiment. FIG. 5 is a diagram illustrating an example of information registered in the query log database according to the embodiment. FIG. 6 is a diagram illustrating an example of information registered in the click log database according to the embodiment. FIG. 7 is a diagram illustrating an example of information registered in the association log database according to the embodiment. FIG. 8 is a diagram illustrating an example of information registered in the search index value database according to the embodiment. FIG. 9 is a diagram illustrating an example of information registered in the selection index value database according to the embodiment. FIG. 10 is a diagram illustrating an example of information registered in the topic degree database according to the embodiment. FIG. 11 is a diagram illustrating an example of processing in which the information distribution apparatus according to the embodiment calculates the topic level. FIG. 12 is a flowchart illustrating an example of the flow of calculation processing executed by the information distribution apparatus according to the embodiment. FIG. 13 is a flowchart for explaining an example of the flow of the associating process executed by the information distribution apparatus according to the embodiment. FIG. 14 is a flowchart for explaining an example of a process flow in which the information distribution apparatus according to the embodiment calculates the topic level. FIG. 15 is a diagram illustrating an example of a hardware configuration of a computer that realizes the function of the information distribution apparatus.

  Hereinafter, a mode for carrying out a calculation device, a calculation method, and a calculation program according to the present application (hereinafter referred to as “embodiment”) will be described in detail with reference to the drawings. Note that the calculation device, the calculation method, and the calculation program according to the present application are not limited by this embodiment. In the following embodiments, the same portions are denoted by the same reference numerals, and redundant description is omitted.

[1. Information distribution device]
First, an example of an information distribution device 10 that is an example of a calculation device will be described with reference to FIG. FIG. 1 is a diagram illustrating an example of an information distribution apparatus according to the embodiment. In FIG. 1, an overview of processing in which the information distribution apparatus 10 distributes target information, which is information on products and services that are targets of electronic commerce such as an EC site, to the user terminal 100 used by the user. Described. In the following description, products and services that are targets of electronic commerce are collectively referred to as transaction targets.

  For example, in the example illustrated in FIG. 1, the information distribution apparatus 10 can communicate with the user terminal 100 via a predetermined network N such as the Internet. The information distribution apparatus 10 can distribute target information to an arbitrary number of user terminals 100.

  The information distribution device 10 manages transaction target information and distributes the transaction target information to the user terminal 100, thereby providing a device for presenting the transaction target to the user, that is, an electronic commerce service. A server on the front end side of the system. For example, the information distribution device 10 may be realized by an information processing device such as a server, or may be realized by a plurality of information processing devices such as a cloud system operating in cooperation. To describe a more specific example, the information distribution apparatus 10 realizes the display process described below using information in a KVS (Key-Value Store) format or information in a HDFS (Hadoop Distributed File System) format. Realized by a group of servers.

  In the following explanation, the transaction target includes provision of various services such as provision of various services, distribution of audio contents, distribution of still images or moving images, and the target information includes information on these services. Shall be. In addition, products and services subject to electronic commerce include products and services traded on an auction site or the like as well as products and services traded on an EC site or the like.

  Further, in the provision of products and services that are subject to electronic commerce, each time a transaction is performed, exchange of consideration may or may not occur. For example, the transaction target described below may include a free moving image distribution service, a moving image distribution service in which a monthly fee is charged, and the like. In the following description, products and services that are transactions may be simply referred to as products.

  The user terminal 100 is a terminal device used by a user, and is realized by, for example, a mobile terminal such as a smartphone, a tablet terminal, or a PDA (Personal Digital Assistant), a desktop PC (Personal Computer), a notebook PC, or the like. Is done.

  For example, when accessing the EC site or the like, the user terminal 100 requests the information distribution apparatus 10 to distribute the target information. In such a case, the information distribution device 10 distributes one or more pieces of target information to the user terminal 100. As a result, the user terminal 100 presents the transaction target to the user by displaying the target information. Then, after accepting the distribution of the target information from the information distribution device 10, the user terminal 100 communicates with other server devices (not shown) to display a web page for purchasing a transaction target or perform a transaction. The target purchase and settlement are executed.

[2. Calculation processing executed by information distribution device]
Here, a plurality of pieces of target information may be displayed in a ranking format on the top page of the EC site. The ranking of such target information depends not only on the newness and price order of the transaction target, but also on how much the transaction target is selling, how much the transaction target is distributed, and how much the transaction target is talking about. In some cases, they are generated in the order of topics, that is, in order of topic level. As described above, when the target information is arranged and displayed in the order of the topic level, the information distribution apparatus 10 can preferentially present the transaction target that is the current topic, so that the user clearly identifies the purchase target. Even if it is not stipulated, it is possible to guide users who are sensitive to fashion to an electronic commerce site.

  However, in the prior art, the priority based on the search history of the transaction object is corrected by the priority based on the result of the web search using the transaction object name as the search query, so the actual popularity is reflected in the priority. You may not be able to. For example, in the prior art, products that are searched or purchased on a daily basis, such as daily necessities, products that have been subjected to SEO (Search Engine Optimization) measures, products that are simply topical and not sold well The priority of increases. For this reason, in the prior art, it may not be possible to preferentially present a product that is actually popular, such as a product that is sold as a result of actual discussion or a product purchased by a user who is an early majority. .

  Therefore, the information distribution apparatus 10 executes the following calculation process, and distributes the ranking of the target information created based on the execution result of the calculation process to the user terminal 100. Specifically, the information distribution apparatus 10 acquires behavior information indicating a history of user behavior regarding selection of a selection target and input information input by the user. And the information delivery apparatus 10 is input including the information related to the predetermined selection object among the action information indicating the history of the user's action related to the selection of the predetermined selection object and the history of the search query. Using the information, a priority when presenting the predetermined selection target is calculated.

[2-1. About selection information)
Here, the selection target is information indicating the target selected by the user. For example, the transaction target itself, the web page for purchasing the transaction target, the link to the web page for purchasing the transaction target, the transaction target This is information indicating information related to a transaction target, such as a store that handles the object, and a target such as an object. For example, the information distribution apparatus 10 acquires information indicating the history of the user's behavior regarding the selection of the transaction target as the behavior information.

  For example, when a web search is performed, a link to a web page for selling a transaction target related to the search query, a web page on which information about the transaction target is posted, and the like are displayed. Therefore, the information distribution apparatus 10 acquires, as action information, a click log that is information indicating the history of the link selected by the user from the links included in the web search result.

  Moreover, the information delivery apparatus 10 specifies the selection object which the user selected using the click log. And the information delivery apparatus 10 links | relates action information and the selection target relevant to the action information. For example, the information distribution apparatus 10 specifies a web page (hereinafter referred to as a link destination) displayed when a link selected by the user is selected, and specifies a selection target related to the specified link destination. To do. More specifically, the information distribution apparatus 10 deals with products and services sold at the specified link destination, products and services whose target information is posted at the link destination, moving images arranged at the link destination, and the like. Identify as a target. And the information delivery apparatus 10 produces | generates the correlation log which linked | related the click log and the transaction object specified from the click log.

[2-2. About input information)
Input information is arbitrary information input by the user. For example, the information distribution apparatus 10 acquires, as input information, a search query history in web search as input information, and uses input information including information indicating a predetermined selection target among the acquired input information. The priority of the selection target is calculated. Note that the information distribution apparatus 10 may acquire information posted to an SNS, a microblog, or the like, that is, information posted on the web by a user as input information.

[2-3. (Calculation of priority)
Here, the purchase history of the transaction target can be an indicator of the willingness of the user to actually select the transaction target. Also, not only users who actually purchased or selected a transaction target, but also users who have transitioned to a web page where the transaction target was posted as a result of a web search are willing to actually purchase that transaction target. It is considered strong. For this reason, the transaction object related to the link destination included in the web search result is predicted to be a transaction object that is highly likely to be actually purchased, that is, a popular transaction object. As a result, the click log associated with a certain transaction target can be an indicator of the willingness of the user to actually select the transaction target, such as the willingness to actually purchase the transaction target.

  In addition to the purchase of a transaction target, a distribution history such as a moving image or an arbitrary selection target can be associated with a history of an action of selecting a link destination related to the selection target, that is, a certain selection target. The action information can be an indicator of a willingness to actually select the selection target. Therefore, the information distribution apparatus 10 describes, for each selection target, a first index (hereinafter referred to as a selection index value) that serves as an index of willingness to select the selection target from action information regarding selection of the selection target. ) Is calculated.

  In addition, information input by the user can be an index of topicality of a selection target related to the information. For example, the name of a transaction target with high topicality is considered to have more web searches, the number of searches on EC sites, auction sites, etc., the number posted on the web, etc. than the name of a transaction target with low topicality. Get it. Therefore, for each selection target, the information distribution apparatus 10 uses a second index (hereinafter referred to as a search index value) as an index of topicality of the selection target from input information including information indicating the selection target. Is calculated.

  And the information delivery apparatus 10 calculates the priority of the selection object for every selection object using the calculated selection index value and search index value. That is, the information distribution device 10 uses the index of the willingness to actually select the selection target and the topicality index of the selection target to determine whether the selection target is actually selected. The priority that indicates the priority, that is, the priority that reflects the actual popularity is calculated. Then, the information distribution device 10 generates a web page in which information indicating a selection target (for example, target information) is arranged in a ranking format based on the calculated priority, and transmits the generated web page to the user terminal 100. To do.

  As described above, the information distribution apparatus 10 acquires the behavior information and the input information, and uses the behavior information related to the predetermined selection target and the input information including the information indicating the predetermined selection target to determine the predetermined selection target. The priority when presenting is calculated. For this reason, the information distribution apparatus 10 can calculate the priority reflecting the actual popularity.

  For example, daily necessities and the like are actually purchased many times, so the value of the selection index value can be high, but since it is not a topic, the value of the search index value is low. On the other hand, a product for which SEO countermeasures have been taken or a product that has just become a topic and is not actually sold has a high search index value and a low selection index value. However, both the search index value and the selection index value of a product that is actually topical and that is actually sold, that is, a product that is actually popular, is high. Therefore, the information distribution apparatus 10 can preferentially present the priority of a product that is actually popular with respect to the early majority, etc., higher than the priority of other products.

[2-4. Example of calculation process executed by information distribution apparatus 10]
Hereinafter, an example of a calculation process executed by the information distribution apparatus 10 will be described with reference to FIG. In the following description, an example of a process for calculating a priority for each transaction target using a transaction target as an electronic commerce target will be described.

  First, the information distribution apparatus 10 acquires behavior information related to an arbitrary transaction target (step S1). More specifically, the information distribution apparatus 10 specifies a history including a selection query that is a search query related to a transaction target from the history of web search. And the information delivery apparatus 10 specifies URL of the link destination which the user selected from the search result when the specified selection query is input, and uses the click log which matched the selection query and URL as action information. get.

  In the example shown in FIG. 1A, the user selects a link to “web page A” from the search result when “selection query # A # B”, which is a search query composed of a plurality of words, is input. There is a web search history indicating the selection. Here, “selection query # A # B” indicates that “selection query #A” and “selection query #B” are input as search queries. In such a case, the information distribution apparatus 10 associates “URL # A”, which is the URL of “Web page A”, with “Selected query # A # B”, as shown in FIG. Get click log.

  Further, the information distribution apparatus 10 acquires a search query history, that is, a query log input from the web search history as input information (step S2). For example, the information distribution apparatus 10 acquires a search query such as “search query #A” or “search query #C”. Note that the information distribution apparatus 10 acquires a search query history including a plurality of words such as “search query # A # B” and “search query # B # C” as a query log.

  In such a case, the information distribution device 10 identifies a transaction target to be selected from the click log, and generates a tied log that associates the transaction target with the behavior information (step S3). For example, the information distribution apparatus 10 specifies a transaction target related to “Web page A” such as a product sold on “Web page A” from “URL #A” included in the click log.

  Here, any technique can be applied to the technique for specifying the transaction target. For example, in a site that conducts electronic commerce or a site that distributes moving images, it is considered that information that identifies a transaction target by a regular expression for each site is included in the URL. Therefore, the information distribution apparatus 10 may specify information for identifying the transaction target from the URL included in the click log, and acquire the transaction target identifier from the specified information. For example, the information distribution apparatus 10 extracts “product ID #A”, which is a transaction target identifier (hereinafter referred to as product ID), from “URL #A”, and extracts the extracted “product ID #A” and “ A tied log in which “selected query # A # B” is associated is generated. The product ID is an identifier of an arbitrary transaction target including not only a so-called product but also a service or the like.

  Moreover, the information delivery apparatus 10 calculates the selection parameter | index value which is a 1st parameter | index from a correlation log for every transaction object (step S4). For example, the information distribution device 10 specifies a link log including “product ID #A”, and serves as an index of willingness to select a transaction target indicated by “product ID #A” from the number of specified link logs. “Selection index value #A” is calculated. Then, the information distribution apparatus 10 stores “product ID #A”, “selection query # A # B”, and “selection index value #A” in association with each other. Note that the selection index value may be calculated by a calculation process similar to a so-called buzz score.

  Subsequently, the information distribution apparatus 10 calculates a search index value that is a second index from the query log for each search query (step S5). For example, the information distribution apparatus 10 calculates “search index value #A” of “search query #A” based on the number of “search query #A” included in the query log. Further, the information distribution apparatus 10 calculates “search index value #B” of “search query #B” based on the number of “search query #B” included in the query log. The search index value may be calculated by a calculation process similar to a buzz score of a so-called search query.

  And the information delivery apparatus 10 calculates the topic degree which is a priority which displays a transaction object from the calculated selection index value and search index value (step S6). For example, as illustrated in FIG. 1C, the information distribution apparatus 10 calculates the topic level of the transaction target indicated by “product ID #A”, and selects “selection” associated with “product ID #A”. “Query #A” and “selection index value #A” are identified.

  Subsequently, the information distribution device 10 specifies a search query including information related to the transaction target indicated by the “product ID #A”. For example, as illustrated in FIG. 1D, the information distribution apparatus 10 identifies the search queries “search query #A” and “search query #B” included in the “selected query # A # B”. Note that the information distribution apparatus 10 further specifies “search query # A # B”, which is the same search query as “selection query # A # B”.

  Then, the information distribution apparatus 10 uses the search index value associated with the identified search query and the selection index value of “product ID #A”, and the topic level of the transaction target indicated by “product ID #A”. "Topic degree #A" is calculated. For example, the information distribution apparatus 10 includes “search index value #A” that is a search index value of “search query #A”, “search index value #B” that is a search index value of “search query #B”, and "Search index value # A # B" that is a search index value of "search query # A # B" and "selection index value #A" that is a selection index value of "product ID #A" are used. Then, “topic level #A”, which is the topic level of the transaction target indicated by “product ID #A”, is calculated. Then, the information distribution apparatus 10 holds “product ID #A” and “topic level #A” in association with each other.

  Thereafter, when the information distribution device 10 receives, for example, a search query for a transaction target from the user terminal 100, the information distribution device 10 specifies a transaction target corresponding to the search query, and sets target information indicating the specified transaction target as a transaction target. Rankings arranged in descending order of topic level are generated (step S7). For example, the information distribution apparatus 10 specifies the transaction target indicated by “product ID #A” and the transaction target indicated by “product ID #B” as the transaction targets corresponding to the search query of the transaction target.

  In such a case, the information distribution device 10 has the topic level “topic level #A” that is the topic level associated with “product ID #A” and the topic level associated with “product ID #B”. “Topic degree #B” is specified, and the target information of the transaction target indicated by “product ID #A” and the transaction indicated by “product ID #B” in the order of the specified “topic level #A” and “topic level #B”. A ranking in which target information indicated by the target is arranged is generated. And the information delivery apparatus 10 transmits the produced | generated ranking to the user terminal 100 (step S8).

  As described above, the information distribution apparatus 10 acquires a click log as behavior information and acquires a query log as input information. And the information delivery apparatus 10 calculates the topic degree of the transaction object using the click log related to the predetermined transaction object in the click log and the query log related to the transaction object in the query log. For this reason, the information delivery apparatus 10 can calculate the topic level that is the priority level reflecting the actual popularity level of the transaction target.

[3. (Connection of information distribution device)
Hereinafter, an example of a functional configuration of the information distribution apparatus 10 that realizes the calculation process illustrated in FIG. 1 will be described. First, an example of a connection relationship between the information distribution apparatus 10 and another apparatus will be described with reference to FIG. FIG. 2 is a diagram illustrating an example of a connection relationship between the information distribution apparatus according to the embodiment and another apparatus.

  As shown in FIG. 2, the information distribution apparatus 10 includes a user terminal 100, a web search log server 110, a shopping search log server 120, an auction search log server 130, an LP (Landing Page) via a network N such as the Internet. The log server 140 is communicably connected.

  The information distribution apparatus 10 is an information processing apparatus that performs the above-described display processing, and is realized by, for example, a single information processing apparatus such as a server apparatus or a plurality of information processing apparatuses such as a cloud system operating in cooperation. Is done.

  The user terminal 100 is an information processing device used by a user who browses various web contents such as web pages. For example, the user terminal 100 is a mobile phone such as a smartphone, a tablet terminal, a PDA (Personal Digital Assistant), a desktop PC (Personal Computer), a notebook PC, or the like.

  For example, the user terminal 100 transmits a transaction target search query to the information distribution apparatus 10 to the information distribution apparatus 10 in accordance with an operation by the user. And the user terminal 100 will display the received ranking, if a ranking is received from the information delivery apparatus 10. FIG.

  The web search log server 110 is an information processing device that collects a history of web searches executed by an arbitrary user, and is realized by a server device, a cloud system, or the like. For example, the web search log server 110 collects a web search log that associates the date and time when the web search was executed, the contents of the search query, and the like.

  The shopping search log server 120 is an information processing device that collects a history of transaction target search processing executed by an arbitrary user at a shopping site such as an EC site, and is realized by a server device, a cloud system, or the like. For example, the shopping search log server 120 collects a shopping search log that associates the date and time when the shopping search was executed, the content of the search query, the transaction target selected by the user from the search result, and the like.

  The auction search log server 130 is an information processing device that collects a history of transaction target search processing executed by an arbitrary user at an auction site, and is realized by a server device, a cloud system, or the like. For example, the auction search log server 130 collects the auction search log by associating the date and time when the auction search was executed, the contents of the search query, the transaction target selected by the user from the search result, and the like.

  The LP log server 140 is an information processing apparatus that collects an LP log indicating a link destination selected by a user from a result of web search or the like, and is realized by a server apparatus, a cloud system, or the like.

  Based on such a connection relationship, the information distribution apparatus 10 executes the following processing. First, the information distribution apparatus 10 acquires a web search log from the web search log server 110. And the information delivery apparatus 10 hold | maintains the acquired web search log as a query log.

  Further, the information distribution apparatus 10 acquires an LP log from the LP log server 140 and specifies a search log corresponding to the acquired LP log. Then, the information distribution device 10 obtains the link destination URL selected by the user in the web search from the acquired LP log and the specified search log, and the search query (that is, the selection query) input in the web search. Generate a click log associated with it, and retain the generated click log.

  In addition, the information delivery apparatus 10 acquires a shopping search log and an auction search log from the shopping search log server 120, the auction search log server 130, etc. by performing the process similar to the process mentioned above. And the information delivery apparatus 10 may produce | generate and hold | maintain a query log and a click log from the acquired shopping search log and auction search log.

[4. Example of functional configuration of information distribution apparatus]
Next, an example of a functional configuration of the information distribution apparatus 10 according to the embodiment will be described with reference to FIG. FIG. 3 is a diagram illustrating an example of a functional configuration of the information distribution apparatus according to the embodiment. As illustrated in FIG. 3, the information distribution apparatus 10 includes a communication unit 20, a storage unit 30, and a control unit 40.

  The communication unit 20 is realized by, for example, a NIC (Network Interface Card). The communication unit 20 is connected to the network N in a wired or wireless manner, and transmits / receives information to / from the user terminal 100, the web search log server 110, the shopping search log server 120, the auction search log server 130, and the LP server 140. Do.

  The storage unit 30 is realized by, for example, a semiconductor memory device such as a RAM (Random Access Memory) or a flash memory, or a storage device such as a hard disk or an optical disk. In addition, the storage unit 30 includes a target information database 31, a query log database 32, a click log database 33, a link log database 34, a search index value database 35, a selection index value database 36, and a topic level database 37 (hereinafter, each of the databases 31 to 31). 37 may be collectively called 37).

  Hereinafter, an example of information registered in each of the databases 31 to 37 stored in the storage unit 30 will be described with reference to FIGS.

  First, an example of information registered in the target information database 31 will be described with reference to FIG. FIG. 4 is a diagram illustrating an example of information registered in the target information database according to the embodiment. In the example shown in FIG. 4, the target information database 31 is registered with target information to be traded and a product ID that is an identifier of the trade target in association with each other.

  For example, in the example illustrated in FIG. 4, the product ID “product ID #A” and the transaction target target information “target information #A” indicated by the product ID “product ID #A” are registered in association with each other. In the example illustrated in FIG. 4, a conceptual value such as “target information #A” is described as an example of the target information, but actually, information such as a description of a transaction target and a price is stored as the target information. .

  Next, an example of information registered in the query log database 32 will be described with reference to FIG. FIG. 5 is a diagram illustrating an example of information registered in the query log database according to the embodiment. In the example illustrated in FIG. 5, a search query and a search date and time are registered in the query log database 32 in association with each other as a query log.

  Here, the search query is information indicating a search query used for web search. The search date and time is information indicating the date and time when the web search was executed. For example, in the example illustrated in FIG. 5, it is registered in the query log database 32 that a web search using the search query “search query #A” has been performed on the search date “2015/01/02/03: 04”. ing. In the query log database 32, other information related to web search may be registered as a query log in association with each other.

  Next, an example of information registered in the click log database 33 will be described with reference to FIG. FIG. 6 is a diagram illustrating an example of information registered in the click log database according to the embodiment. In the example illustrated in FIG. 6, a selection query, a click URL, and an operation date / time are associated and registered as a click log in the click log database 33.

  Here, the selection query is a search query in a web search or the like in which a user selects a link destination from a search result. The click URL is a link destination URL selected from the search result by the user. The operation date and time is the date and time when the user selects a link destination, that is, the date and time when the user performs an operation of selecting a selection target.

  For example, in the example illustrated in FIG. 6, the URL of the link destination selected by the user from the result of the web search using the selection query “selection query # A # B” is “URL # A” in the click log database 33. Yes, the fact that the operation date and time when the link destination is selected is “2015/01/02/04: 04” is registered. In the click log database 33, various types of information related to an operation for the user to select a selection target may be further associated and registered as a click log.

  Next, an example of information registered in the association log database 34 will be described with reference to FIG. FIG. 7 is a diagram illustrating an example of information registered in the association log database 34 according to the embodiment. In the example illustrated in FIG. 7, a selection query, a product ID, and a date are registered in association with each other in the association log database 34. Here, the product ID is an identifier indicating a transaction object related to the associated selection query. The date is the operation date and time included in each click log including the associated selection query.

  For example, in the example illustrated in FIG. 7, the product ID of the transaction target related to the selection query “selection query # A # B” is “product ID # A” in the association log database 34, and the selection query “selection query #” This indicates that the web search using “A-1” was executed at the date “2015/01/02/04: 04”. In the association log database 34, various types of information related to the operation of the user who selects the link destination associated with the transaction target indicated by the product ID when the selection query is input are further registered in association with each other. It may be. In the association log database 34, a cookie corresponding to the user who executed the web search may be registered.

  Next, an example of information registered in the search index value database 35 will be described with reference to FIG. FIG. 8 is a diagram illustrating an example of information registered in the search index value database according to the embodiment. As shown in FIG. 8, in the search index value database 35, a search query and a search index value of a transaction target related to the search query are registered in association with each other. For example, in the example illustrated in FIG. 8, the search query “search query #A” corresponds to the search index value “search index value #A” indicating the topicality of the transaction target related to the search query “search query #A”. It is registered with it.

  Next, an example of information registered in the selection index value database 36 will be described with reference to FIG. FIG. 9 is a diagram illustrating an example of information registered in the selection index value database according to the embodiment. As shown in FIG. 9, in the selection index value database 36, selection queries, product IDs, and transaction target selection index values indicated by the product IDs are registered in association with each other. For example, in the example illustrated in FIG. 9, the selection query “selection query # A # B”, the product ID “product ID # A”, and the selection indicating the willingness to select the transaction target indicated by the product ID “product ID # A”. The index value “selected index value # A # B” is registered in association with each other.

  Next, an example of information registered in the topic level database 37 will be described with reference to FIG. FIG. 10 is a diagram illustrating an example of information registered in the topic degree database according to the embodiment. As shown in FIG. 10, in the topic level database 37, the product ID and the topic level calculated for the transaction target indicated by the product ID are registered in association with each other. For example, in the example illustrated in FIG. 10, the product ID “product ID #A” and the topic level “topic level #A” calculated for the transaction target indicated by “product ID #A” are registered in association with each other. .

  Returning to FIG. 3, the description of the functional configuration of the information distribution apparatus 10 will be continued. The control unit 40 is stored in a storage device inside the information distribution apparatus 10 by, for example, a CPU (Central Processing Unit), an MPU (Micro Processing Unit), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or the like. The various programs are executed by using the RAM as a work area.

  As shown in FIG. 3, the control unit 40 includes a log acquisition unit 41, an association unit 42, a search index value calculation unit 43, a selection index value calculation unit 44, a topic level calculation unit 45, and a product information distribution unit 46.

  The log acquisition unit 41 acquires behavior information indicating a history of user behavior related to selection of a selection target, and input information that is information input by the user. For example, the log acquisition unit 41 acquires a web search log, a shopping search log, an auction search log, and an LP log. Then, the log acquisition unit 41 generates a query log from the acquired various logs, and registers the generated query log in the query log database 32.

  In addition, the log acquisition unit 41 acquires behavior information indicating a history of user behavior related to selection of a transaction target to be a target of electronic commerce from the acquired various logs. For example, the log acquisition unit 41 selects the selected query used for the web search, the URL of the link destination selected by the user (that is, the click URL), and the date and time when the web search was performed (that is, from the acquired various logs). , Operation date and time). That is, the log acquisition unit 41 is a click log indicating a history of a selection target selected by a user from a selection target included in a result of a web search for a set of a selection query and a link destination URL selected by the user. Get as. Then, the log acquisition unit 41 registers the click log in the click log database 33.

  The associating unit 42 generates an associating log that associates the behavior information indicating the history of the user's behavior related to the selection of the predetermined selection target in the behavior information and the predetermined selection target. Specifically, the associating unit 42 specifies the transaction target that is the target of the electronic commerce in the linked web page selected by the user, and selects the identified transaction target and the link to the web page. A link log is generated that links the action information indicating the. More specifically, the associating unit 42 specifies a selection target related to the selected linked web page, and includes the selected selection target and action information indicating an operation for selecting a link to the web page. Tie it.

  For example, the associating unit 42 refers to the click log database 33 and executes the following processing. First, the associating unit 42 specifies a transaction target sold on the web page indicated by the click URL. A more specific example will be described. The associating unit 42 identifies a service that delivers the web page indicated by the click URL, and in the identified service, the range corresponding to the product ID is included in the click URL. To identify Then, the associating unit 42 converts the information in the specified range in the click URL according to the pattern of the normalized expression that is different for each service, so that the product ID of the transaction target related to the web page indicated by the click URL Is identified.

  Then, the association unit 42 converts the click URL of each click log into the specified product ID, and registers the click log after conversion in the association log database 34 as an association log. In addition, the associating unit 42 indicates the number of click logs (that is, the user indicated by the cookie indicates the product ID for each set of the user cookie, the product ID, and the selection query included in each click log. The number of clicks of the transaction target) may be specified as the number of clicks, and the specified number of clicks, the product ID, and the set of selection queries may be registered as an association log.

  The search index value calculation unit 43 calculates a search index value from the query log. For example, the search index value calculation unit 43 acquires a query log from the query log database 32. Then, the search index value calculation unit 43 calculates a search index value based on the number of query logs for each search query. To explain a more specific example, the search index value calculation unit 43 calculates the number of search queries every two weeks (that is, the number of searches) in time series for each search query, Correlation analysis is performed with the number of searches, and a correlation coefficient is calculated. Then, the search index value calculation unit 43 registers the calculated correlation coefficient as a search index value in the search index value database 35 in association with the search query.

  The selection index value calculation unit 44 calculates a selection index value from the click log. For example, the selection index value calculation unit 44 calculates the number of click logs (that is, the number of clicks) every two weeks in time-series order from the click log database 33 for each combination of the selection query and the product ID. Correlation analysis is performed with the number of searches serving as a value calculation base, and a correlation coefficient is calculated. Then, the selection index value calculation unit 44 registers the calculated correlation coefficient as a selection index value in the selection index value database 36 in association with the combination of the selection query and the product ID.

  Hereinafter, the topic degree calculation unit 45 includes behavior information indicating a history of user behavior regarding selection of a predetermined selection target in the behavior information, and input information including information regarding the predetermined selection target among the input information. Use to calculate the priority when presenting a predetermined selection target. More specifically, the topic level calculation unit 45 includes behavior information indicating a history of the user's behavior related to selection of a predetermined transaction target, and input information including information related to the predetermined transaction target, among the behavior information. Is used to calculate the priority when presenting a predetermined transaction object.

  That is, the topic level calculation unit 45 calculates the priority of a predetermined selection target using input information including information related to the predetermined selection target and behavior information associated with the predetermined selection target. In other words, the topic level calculation unit 45 calculates the priority of the predetermined selection target using the behavior information associated with the predetermined selection target and the input information including the selection query included in the behavior information. More specifically, the topic level calculation unit 45 calculates the priority of the predetermined selection target using the search index value and the selection index value related to the predetermined selection target.

  Hereinafter, an example of processing executed by the topic level calculation unit 45 will be described in detail. First, the topic level calculation unit 45 performs matching between the search query registered in the search index value database 35 and the selection query registered in the selection index value database 36. For example, the topic level calculation unit 45 extracts one set of a selection query and a product ID registered in the selection index value database 36 as a processing target. Next, the topic level calculation unit 45 extracts the search index value of the search query related to the selected query included in the extracted set from the search index value database 35.

  For example, the topic level calculation unit 45 extracts a set of “selection query # A # B” and product ID “product ID # A”. In such a case, the topic level calculation unit 45 uses the same search query as “selection query # A # B” or “search query # A # B” as a search query related to “selection query # A # B”. Identify all search queries included in.

  For example, “selection query #A” is a word indicating the name of a movie, and “selection query #B” is a word indicating the type of transaction object (for example, “DVD (Digital Video Disc)”, “mug cup” It is a word indicating goods etc.). In such a case, the topic level calculation unit 45 specifies “search query #A” that is a search query including the name of a movie and “search query #B” that is a search query indicating the type of goods. Note that the topic level selection unit 45 may also specify a search query related to “selection query #A” or “selection query #B” (for example, the name of an actor who appears in a movie).

  Then, the topic level calculation unit 45 extracts all search index values associated with the specified search query from the search index value database 35. That is, the topic level calculation unit 45 identifies a search query including information related to a transaction target associated with “selected query # A # B”, and selects all selection indexes associated with the identified search query. Extract the value. For example, in the example illustrated in FIG. 8, the topic level calculation unit 45 searches the search index values “search index value #A” and “search query #B” associated with “search query #A”. An index value “search index value #B”, a search index value “search index value # A # B” associated with “search query # A # B”, and the like are extracted.

  Subsequently, the topic level calculation unit 45 calculates an average value of the extracted search index values. Then, the topic level calculation unit 45 considers a predetermined parameter, and sums the selection index value associated with the set to be processed and the average value of the extracted search index values (hereinafter referred to as index sum). ) Is calculated. For example, the topic level calculation unit 45 calculates a value obtained by adding a predetermined parameter α to an average value of search index values and a value obtained by adding a predetermined parameter β to a selection index value associated with a set to be processed. The sum is calculated as the index sum. Here, it is assumed that the values of α and β are values at which the sum of the values is 1.

  Further, the topic level calculation unit 45 calculates the above-described index sum value for each set of all selection queries and product IDs. Then, the topic level calculation unit 45 calculates the priority for each product ID based on the distribution of the calculated index sum value. For example, the topic level calculation unit 45 calculates the standard deviation of the index sum values, assumes that the distribution of the index sum values follows a normal distribution, and to which region of the distribution each index sum value belongs. The threshold value for determining is calculated. For example, the topic level calculation unit 45 has a value of each index sum belonging to any of a range of less than 80%, a range of 80% to 90%, a range of 90% to 96%, or a range of 96% or more. Thus, the threshold value is calculated. The threshold value is the sum of the average value of the index sum and the value obtained by multiplying the standard deviation of the index sum value by a predetermined coefficient. By changing the coefficient value, A threshold value as a boundary can be calculated.

  Next, the topic level calculation unit 45 calculates a score in accordance with the range to which each index sum belongs. First, the topic level calculation unit 45 sets a score range condition according to a range to which each index sum belongs. For example, the topic level calculation unit 45 sets the score value calculated from the index sum belonging to the range less than 80% to 50 to 60, and the score value calculated from the index sum belonging to the range from 80% to 90% Is 65 to 79. Further, the topic level calculation unit 45 sets the score value calculated from the index sum belonging to the range of 90% to 96% to 80 to 84, and the score value calculated from the index sum belonging to the range of 96% or more. Is 95-100.

  Then, the topic level calculation unit 45 learns the values of a and b of the linear function (affine function) y = ax + b so that the above-described conditions are satisfied, where y is the priority value and x is the index sum value. . That is, the topic level calculation unit 45 learns a model that calculates the topic level from the sum of the search index value and the selection index value using the distribution of the search index value and the selection index value as a feature. Then, the topic level calculation unit 45 calculates the topic level for each transaction target using the learned model, and registers the calculated topic level and the product ID of the transaction target in the topic level database 37 in association with each other.

  Depending on the user's behavior, a plurality of selection queries may be associated with one selection target. In such a case, the topic level calculation unit 45 calculates all the priorities for each set of the predetermined selection target and the search query, and based on the calculated priority values, the priority of the predetermined selection target May be calculated. For example, the topic level calculation unit 45 specifies all combinations of selection queries and product IDs including a certain product ID, calculates a priority for each specified set, and calculates the largest value among the calculated priorities It may be the topic level of the selection target indicated by the product ID.

  The merchandise information distribution unit 46 distributes merchandise information. For example, when the product information distribution unit 46 receives a search query from the user terminal 100, the product information distribution unit 46 identifies the product ID of the transaction target corresponding to the received search query, and sets the topic level associated with the identified product ID as the topic level. It is specified from the database 37. And the merchandise information distribution part 46 produces | generates the ranking which arranged the merchandise information of transaction object corresponding to the received search query in an order from the one with the high value of the specified topic degree, and makes the produced | generated ranking to the user terminal 100. Send.

[5. (About query support)
Next, an example of processing for calculating the topic level will be described with reference to FIG. FIG. 11 is a diagram illustrating an example of processing in which the information distribution apparatus according to the embodiment calculates the topic level. In the example illustrated in FIG. 11, an example of processing for calculating the topic level of the transaction target “product #A” associated with the word “character #A” will be described. For example, the user performs a web search using “character #A” as a search query Q01. In such a case, the user terminal 100 displays the search result W01 in which links L01 to L03 to the web page related to “character #A” are arranged.

  When the search result W01 is displayed in this way, for example, when the user clicks the position of the cursor C01 along the link L01, that is, when the user selects the link L01, the information distribution apparatus 10 A click log in which the search query Q01 is associated with the link destination URL of the link L01 is acquired. For example, as illustrated in FIG. 11A, the information distribution apparatus 10 acquires a click log in which “character #A” and “URL #A” are associated with each other. In such a case, as shown in FIG. 11B, the information distribution apparatus 10 identifies the product “product #A” related to the web page indicated by “URL #A”, and identifies the identified “product #A ”And“ Character #A ”are generated. In the example shown in FIG. 11, it is assumed that the link L01 having “URL #A” as the link destination is selected “1000 times” from the search result using “character #A” as the selection query.

  On the other hand, the information distribution apparatus 10 acquires a history Q02 of a search query such as a web search. In such a case, the information distribution apparatus 10 counts the number of times each search query has been searched. For example, as shown in FIG. 11C, the information distribution apparatus 10 counts the number of times “2000 times” the search query “character #A” has been searched.

  Here, in consideration of the search for a transaction object, when searching for a medium in which a predetermined movie is recorded, the user inputs not only the name of the movie but also the name of the character appearing in the movie as a search query, There is a case where a web page for selling a movie recording medium is selected from a search result. Also, for example, the user may select a web page that sells a movie recording medium or a web page that sells movie related goods from a search result that uses the name of a movie as a search query. is there. On the other hand, considering the contents of the web search, when a movie is popular, the user not only searches the web using the name of the movie, but also uses the names of characters, etc. that appear in the movie. Web search.

  For this reason, when the information distribution apparatus 10 matches the click log and the query log using the trade name of the transaction target, the information distribution apparatus 10 may not reflect the actual popularity of the transaction target in the topic level. For example, when a query log including a search query related to a trade target product name is a matching target in the query log, the information distribution device 10 is not related to the product name but is related to the popularity of the trade target. The query log containing is excluded from the matching target.

  Therefore, as shown in FIG. 11D, the information distribution apparatus 10 performs matching between the click log selection query associated with the transaction target and the search query included in the query log. For example, in the example illustrated in FIG. 11, the information distribution apparatus 10 matches the click log including the selection query “character #A” with the query log including the search query “character #A” in FIG. ), The degree of topic of “product #A” as a transaction target is calculated in consideration of the number of searches of the search query “character #A”.

  That is, the information distribution apparatus 10 calculates the topic level of the product related to the selection query in consideration of the same search query as the selection query and the number of searches of all the search queries included in the selection query. As a result, the information distribution apparatus 10 considers not only the name of “product #A” but also the number of searches of the search query related to “product #A”, and the topic level “topic level #A” of “product #A”. Therefore, the topic level reflecting the actual popularity can be calculated. For example, when a movie or a performer of a movie is a topic, the information distribution device 10 can increase the topic level of goods or the like related to the movie.

[6. Procedure of information executed by information distribution apparatus]
Next, the flow of processing executed by the information distribution apparatus 10 will be described with reference to FIGS. First, the flow of calculation processing executed by the information distribution apparatus 10 will be described with reference to FIG. FIG. 12 is a flowchart illustrating an example of the flow of calculation processing executed by the information distribution apparatus according to the embodiment.

  For example, the information distribution apparatus 10 acquires a query log and a click log (step S101). Subsequently, the information distribution apparatus 10 calculates a search index value from the query log (step S102). Subsequently, the information distribution apparatus 10 executes a linking process for linking the click log and the transaction target (step S103), and calculates a selection index value from the linking log (step S104). And the information delivery apparatus 10 calculates the topic level of a transaction object using a search parameter | index value and a selection parameter | index value (step S105), and complete | finishes a process.

  Next, the flow of the pegging process shown in step S103 in FIG. 12 will be described using FIG. FIG. 13 is a flowchart for explaining an example of the flow of the associating process executed by the information distribution apparatus according to the embodiment.

  For example, the information distribution apparatus 10 specifies the click URL associated with the selection query from the click log (step S201). Subsequently, the information distribution apparatus 10 extracts a character string indicating a transaction target from the specified URL (step S202). Then, the information distribution apparatus 10 converts the extracted character string into a product ID indicating a transaction target according to a rule corresponding to the site (step S203), and generates a linked log in which the selection query and the product ID are associated with each other. (Step S204), the process ends.

  Next, the flow of processing for calculating the topic level shown in step S105 in FIG. 12 will be described with reference to FIG. FIG. 14 is a flowchart for explaining an example of a process flow in which the information distribution apparatus according to the embodiment calculates the topic level.

  For example, the information distribution apparatus 10 extracts a search query associated with a certain selection query, that is, a search index value associated with all search queries included in the selection query or the same search query as the selection query (step S301), an average value of the extracted search index values is calculated (step S302). Subsequently, the information distribution apparatus 10 calculates the sum of the value obtained by integrating the predetermined parameter to the selected index value and the value obtained by integrating the predetermined parameter to the average value as the index sum (step S303).

  And the information delivery apparatus 10 determines whether the index sum was calculated about all the transaction objects, and when not calculating (step S304: No), performs step S301. On the other hand, if the information distribution device 10 has calculated the index sum for all the transaction targets (step S304: Yes), the threshold for calculating the topic level based on the standard deviation of the calculated index sum, that is, each A threshold value for dividing the region to which the index sum belongs is calculated (step S305). Further, the information distribution apparatus 10 learns a model for calculating the topic level according to the range divided by the threshold (step S306). And the information delivery apparatus 10 calculates the topic degree of each transaction object from each index sum using a model (step S307), and complete | finishes a process.

[7. (Modification)
The information distribution apparatus 10 according to the above-described embodiment may be implemented in various different forms other than the above-described embodiment. Therefore, in the following, another embodiment of the information distribution apparatus 10 will be described.

[7-1. (Processing considering the characteristics of the selection target)
For example, in various searches, products to be selected ("skirt", "one piece", "tote bag", such as "design A skirt", "design B one piece", "design C tote bag", etc. ”) And characteristics of the product (“ Design A ”,“ Design B, “Design C”, etc.) ”may be input as a search query. When such a search is performed, it is better that the priority is calculated for each set of a product and characteristics of the product.

  Therefore, the information distribution apparatus 10 may calculate the priority such as the topic level in consideration of the characteristics of the selection target. For example, the information distribution device 10 calculates the priority for each set of the selection target and the characteristics of the selection target using action information related to selection of the selection target and input information including information related to the set. May be performed.

  For example, the information distribution apparatus 10 acquires a click log in which a URL selected by a user is associated with a set of a selection query related to a selection target and characteristics of the selection target. In such a case, the information distribution apparatus 10 associates the URL and a selection target corresponding to the acquired click log. That is, the information distribution apparatus 10 associates a corresponding selection target with a combination of a selection query and a characteristic.

  Then, the information distribution apparatus 10 identifies a search query related to the combination of the selection query and the characteristic, and a query log including the search query related to the selection query included in the combination, the acquired click log, the identified query log, May be used to calculate the priority for each set of selection query and characteristic. By executing such processing, the information distribution apparatus 10 can calculate the priority in consideration of the degree of popularity for each specification of design, color, etc. even for the same product.

[7-2. About input information and action information)
In addition, as long as the information distribution apparatus 10 is a log indicating the behavior of the user who selects the selection target, the information distribution device 10 includes not only information indicating arbitrary behavior on the web but also information indicating arbitrary behavior of the user in the real world. You may employ | adopt as action information. For example, the information distribution device 10 generates not only a web search click log but also a linked log from, for example, a click log for a transaction target search on an auction site or a click log for a transaction target search on an EC site or the like. Also good.

  In addition, for example, the information distribution apparatus 10 acquires, as behavior information, a credit card usage history, a transaction history purchased on an EC site, and a purchase history of a selection target, and a predetermined selection of the acquired behavior information. You may use action information relevant to an object for calculation of a priority of a predetermined selection object. For example, the information distribution apparatus 10 may correct the value of the selection index value corresponding to the selection target using the purchase history of the selection target. By executing such processing, the information distribution apparatus 10 calculates the priority of the selection target based on the purchase history of the selection target in any electronic commerce site or actual store. Priority can be calculated accurately considering the popularity of.

  Moreover, the information delivery apparatus 10 is good also considering the information posted on SNS, a microblog, etc., ie, the information regarding the predetermined selection object posted on the web, as input information. The information distribution apparatus 10 may correct the search index value of the input information based on an index such as how much tweets are tweeted by real-time analysis of tweets.

  For example, the information distribution apparatus 10 can increase the priority of the selection target that is predicted to be popular by correcting the priority of the selection target based on a user's post called an influencer. The information distribution apparatus 10 may cluster users who post information to the SNS according to a predetermined rule, and correct the search index value of the search query according to the number of posts in each cluster.

[7-3. (Processing considering location information)
The information distribution apparatus 10 may calculate the priority in consideration of the position where the web search or the like is performed. For example, the information distribution apparatus 10 acquires a set of behavior information and a position where the behavior indicated by the behavior information is performed. Then, the information distribution device 10 is based on the position where each action is performed based on the history of the user's action related to the selection of the predetermined selection target and the search query candidates including information indicating the predetermined selection target. The priority may be calculated in consideration of weighting. For example, when the search query related to the predetermined selection target is a search query for web search performed at a store that handles the predetermined selection target, the information distribution device 10 determines the value of the search index value of the search query. May be corrected to a higher value.

  The information distribution apparatus 10 associates position information with a store and tag information indicating a selection target category handled by the store in advance, and uses the position information associated with the store in the vicinity of the store. Collect search queries for web searches done. In addition, the information distribution apparatus 10 identifies a popular search query in the category, assuming that the collected search query is a search query belonging to the category indicated by the tag information associated with the store. Then, the information distribution apparatus 10 may calculate the priority of the selection target related to the category using the search index value of the specified search query.

[7-4. (Applicable calculation process)
Note that the above-described calculation processing can be applied not only to products and services that are subject to electronic commerce, but also to any digital content such as products, services, and news that are to be purchased in actual stores. In addition, the information distribution apparatus 10 may calculate priority by selecting a target that does not require exchange of consideration, such as a moving image distributed by a free video distribution service, a free sample, and a monitor recruitment for various services. Good. In addition, the information distribution apparatus 10 may calculate the priority with respect to various services for which consideration is generated in the monthly installment method. In addition, the information distribution apparatus 10 may generate a ranking based on the priority calculated for each transaction target as a search result of the transaction target on the EC site or the auction site.

[7-5. (About search queries that include information related to selection queries)
In the calculation process described above, the information distribution apparatus 10 specifies a search query including information related to the selection query, and calculates the priority of the transaction target corresponding to the selection query using the search index value of the specified search query. did. For example, the information distribution measure 10 specifies the same search query as the selection query and all the search queries included in the selection query as the search query including information related to the selection query. However, the embodiment is not limited to this.

  For example, when a certain actor is talking about the information distribution apparatus 10, it is considered that the information distribution apparatus 10 may inform the priority of a movie in which the actor has appeared. Therefore, the information distribution apparatus 10 searches for a search query including the name of an actor who has appeared in the movie when the transaction name associated with the selection query is a movie recording medium. The priority of the recording medium of the movie may be calculated using the index value. That is, the information distribution apparatus 10 may calculate the priority using an arbitrary search query including information related to the selection query. At least, if a certain search query can suggest a transaction object corresponding to a certain selection query, the information distribution apparatus 10 calculates the priority of the transaction object using the search index value of the search query. May be.

[7-6. Hardware configuration)
Further, the information distribution apparatus 10 according to the embodiment described above is realized by a computer 1000 having a configuration as shown in FIG. 15, for example. FIG. 15 is a diagram illustrating an example of a hardware configuration of a computer that realizes the function of the information distribution apparatus. The computer 1000 includes a CPU 1100, RAM 1200, ROM 1300, HDD 1400, communication interface (I / F) 1500, input / output interface (I / F) 1600, and media interface (I / F) 1700.

  The CPU 1100 operates based on a program stored in the ROM 1300 or the HDD 1400 and controls each unit. The ROM 1300 stores a boot program executed by the CPU 1100 when the computer 1000 is started up, a program depending on the hardware of the computer 1000, and the like.

  The HDD 1400 stores a program executed by the CPU 1100, data used by the program, and the like. The communication interface 1500 receives data from other devices via the communication network 500 (corresponding to the network N) and sends the data to the CPU 1100, and transmits data generated by the CPU 1100 to other devices via the communication network 500. To do. The CPU 1100 controls an output device such as a display and a printer and an input device such as a keyboard and a mouse via the input / output interface 1600. The CPU 1100 acquires data from the input device via the input / output interface 1600. Further, the CPU 1100 outputs the data generated via the input / output interface 1600 to the output device.

  The media interface 1700 reads a program or data stored in the recording medium 1800 and provides it to the CPU 1100 via the RAM 1200. The CPU 1100 loads such a program from the recording medium 1800 onto the RAM 1200 via the media interface 1700, and executes the loaded program. The recording medium 1800 is, for example, an optical recording medium such as a DVD (Digital Versatile Disc) or PD (Phase change rewritable disk), a magneto-optical recording medium such as an MO (Magneto-Optical disk), a tape medium, a magnetic recording medium, or a semiconductor memory. Etc.

  For example, when the computer 1000 functions as the information distribution apparatus 10, the CPU 1100 of the computer 1000 implements the function of the control unit 40 by executing a program loaded on the RAM 1200. In addition, each data in the storage unit 30 is stored in the HDD 1400. The CPU 1100 of the computer 1000 reads and executes these programs from the recording medium 1800. However, as another example, these programs may be acquired from other devices via a communication network.

[8. effect〕
As described above, the information distribution apparatus 10 acquires behavior information indicating a history of user behavior regarding selection of a selection target and input information that is information input by the user. The information distribution device 10 includes behavior information indicating a history of the user's behavior regarding selection of a predetermined selection target among the behavior information, and input information including information related to the predetermined selection target among the input information, Is used to calculate the priority when presenting a predetermined selection target. For this reason, the information delivery apparatus 10 can reflect an actual popularity in the priority at the time of presenting a selection target.

  Moreover, the information delivery apparatus 10 acquires the action information which shows the log | history of the action of the user regarding selection of the transaction object used as the object of electronic commerce as a selection object. And the information delivery apparatus 10 is predetermined | prescribed using the action information which shows the log | history of the action of the user regarding selection of a predetermined transaction object, and the input information containing the information relevant to a predetermined transaction object among action information. Calculate the priority when presenting the transaction object. For this reason, the information delivery apparatus 10 can reflect an actual popularity degree with respect to the priority of the goods and service used as a transaction object in an online shopping mall etc., for example.

  Moreover, the information delivery apparatus 10 acquires the information which shows the log | history of the selection target which the user selected from the selection targets contained in the result of web search as action information. For example, the information distribution apparatus 10 acquires the click log described above. For this reason, the information delivery apparatus 10 can calculate the priority of the selection target based on the action of the user selecting the selection target as a result of the web search.

  Moreover, the information delivery apparatus 10 links | relates the action information which shows the log | history of the action of the user regarding selection of a predetermined | prescribed selection object, and a predetermined selection object among action information. And the information delivery apparatus 10 calculates the priority of a predetermined | prescribed selection object using the input information containing the information relevant to a predetermined | prescribed selection object, and the action information linked | related with the predetermined selection object.

  For example, the information distribution apparatus 10 acquires information indicating the link selected by the user among the links included in the web search result as the behavior information. Then, the information distribution apparatus 10 specifies a selection target related to the web page indicated by the selected link, and associates the specified selection target with behavior information indicating an operation for selecting a link to the web page.

  In addition, for example, the information distribution apparatus 10 identifies a transaction target that is an object of electronic commerce in the web page indicated by the selected link, and behavior information indicating an operation for selecting the identified transaction target and a link to the web page Is linked. And the information delivery apparatus 10 calculates the priority of a predetermined | prescribed transaction object using the input information containing the information relevant to a predetermined | prescribed transaction object, and the action information linked | related with the predetermined transaction object.

  In addition, for example, the information distribution apparatus 10 includes, as the behavior information, information indicating a selected link among a selection query that is a search query used for web search and a link included in a result of the web search using the selection query. Get a pair with. Then, the information distribution device 10 calculates the priority of the predetermined selection target using the behavior information associated with the predetermined selection target and the input information including information related to the selection query included in the behavior information. .

  As described above, the information distribution apparatus 10 calculates the priority of the selection target using the action information associated with the selection target, and thus the actual popularity of the selection target can be reflected in the priority.

  In addition, the information distribution apparatus 10 relates to the action information indicating the history of the user's action related to the selection of the predetermined selection target for each set of the predetermined selection target and the characteristic of the predetermined selection target, and the set. The priority is calculated using the input information including the information. Thus, since the information distribution apparatus 10 calculates the priority for each pair of the selection target and the characteristic, for example, the commodity information of each commodity is presented with a different priority for each characteristic of the commodity. Can do. For example, even if the information distribution apparatus 10 is the same product, it can be presented with different priorities according to its color, design, and the like.

  Further, the information distribution device 10 calculates a selection index value that is an index of a user's willingness to select a predetermined selection target from the behavior information, and performs a predetermined selection from input information including information indicating the predetermined selection target. A search index value that is an index of subject topicality is calculated, and a priority of a predetermined selection target is calculated using the selection index value and the search index value. Further, the information distribution apparatus 10 calculates and calculates an index sum that is a sum of a value obtained by adding a predetermined parameter to a selection index value and a value obtained by adding a predetermined parameter to a search index value for each selection target. The priority for each selection target is calculated based on the distribution of the index sum values. For this reason, the information distribution apparatus 10 can calculate the priority of the selection target with high accuracy from the behavior information and the search history.

  Further, the information distribution apparatus 10 learns a model for calculating the priority from the index sum value using the calculated distribution of the index sum as a feature, and calculates the priority for each selection target using the learned model. To do. For this reason, the information delivery apparatus 10 can calculate the priority of each selection object appropriately.

  Further, the information distribution apparatus 10 calculates the priority of the predetermined selection target using the history of the search query including information related to the predetermined selection target out of the search query history in the web search as the input information. . For this reason, the information delivery apparatus 10 can reflect the latest popularity degree in the priority of selection object.

  In addition, the information distribution apparatus 10 calculates the priority of the predetermined selection target based on information regarding the predetermined selection target posted on the web as input information. For this reason, the information delivery apparatus 10 can appropriately calculate the priority of the selection target that is popular in, for example, SNS and microblogging.

  Moreover, the information delivery apparatus 10 calculates the priority of a predetermined selection object using the purchase history of a predetermined selection object as action information. For this reason, the information distribution apparatus 10 can calculate the priority of the selection target based on the purchase history in the actual store as well as in the electronic commerce, so that the actual popularity can be reflected in the priority. it can.

  In addition, the information distribution apparatus 10 acquires a set of behavior information and a position where the behavior indicated by the behavior information is performed. Then, the information distribution device 10 indicates the behavior information based on the behavior information indicating the history of the user's behavior related to the selection of the predetermined selection target and the search query candidates including information related to the predetermined selection target. The priority of the selection target is calculated in consideration of the weighting according to the position where the action is performed. Thus, since the information distribution apparatus 10 calculates the priority according to the position where the operation for selecting the selection target is performed, the actual popularity can be reflected in the priority.

  As described above, some of the embodiments of the present application have been described in detail based on the drawings. It is possible to implement the present invention in other forms with improvements.

  Moreover, the above-mentioned “section (module, unit)” can be read as “means”, “circuit”, and the like. For example, the control unit can be read as control means or a control circuit.

DESCRIPTION OF SYMBOLS 10 Information delivery apparatus 20 Communication part 30 Storage part 31 Target information database 32 Query log database 33 Click log database 34 Linked log database 35 Search index value database 36 Selection index value database 37 Topic degree database 40 Control part 41 Log acquisition part 42 Linking part 43 Search index value calculation unit 44 Selection index value calculation unit 45 Topic degree calculation unit 46 Product information distribution unit 100 User terminal 110 Web search log server 120 Shopping search log server 130 Auction search log server 140 LP log server

Claims (17)

An acquisition unit that acquires behavior information indicating a history of the user's behavior related to selection of a selection target and input information that is information input by the user;
Among the behavior information, using behavior information indicating a history of user behavior related to selection of a predetermined selection target, and input information including information related to the predetermined selection target among the input information, And a calculation unit that calculates a priority when presenting a predetermined selection target. The acquisition unit acquires, as the selection target, behavior information indicating a history of a user's behavior related to selection of a transaction target to be a target of electronic commerce,
The calculation unit uses the behavior information indicating the history of the user's behavior related to selection of a predetermined transaction target and the input information including information related to the predetermined transaction target in the behavior information. The calculation device according to claim 1, wherein a priority level when presenting a transaction object is calculated. The said acquisition part acquires the information which shows the log | history of the selection target which the user selected from the selection targets contained in the result of web search as said action information. Calculation device. Among the behavior information acquired by the acquisition unit, the behavior information indicating the history of the user's behavior related to the selection of a predetermined selection target, and a linking unit that links the predetermined selection target,
The calculation unit calculates a priority of the predetermined selection target using input information including information related to the predetermined selection target and behavior information associated with the predetermined selection target. The calculation device according to any one of claims 1 to 3. The acquisition unit acquires, as the behavior information, information indicating a link selected by the user among links included in a web search result,
The associating unit identifies a selection target related to the web page indicated by the selected link, and associates the identified selection target with action information indicating an operation for selecting a link to the web page. The calculation device according to claim 4. The associating unit identifies a transaction target that is an object of electronic commerce in the web page indicated by the selected link, and associates the identified transaction target with behavior information indicating an operation for selecting a link to the web page. And
The calculation unit calculates the priority of the predetermined transaction object using input information including information related to the predetermined transaction object and behavior information associated with the predetermined transaction object. The calculation device according to claim 5. The acquisition unit includes, as the behavior information, a selection query that is a search query used for the web search, and information indicating a link selected from links included in a web search result using the selection query. Get a pair,
The calculation unit calculates the priority of the predetermined selection target using behavior information associated with the predetermined selection target and input information including information related to a selection query included in the behavior information. The calculation device according to claim 5 or 6, wherein For each set of the predetermined selection target and the characteristics of the predetermined selection target, the calculation unit relates to behavior information indicating a history of user behavior related to selection of the predetermined selection target and the set The calculation device according to any one of claims 1 to 7, wherein the priority is calculated using input information including information. The calculation unit calculates a first index serving as a motivation index for a user to select the predetermined selection target from the behavior information, and inputs the predetermined index from input information including information indicating the predetermined selection target Calculating a second index that is an index of topicality of the selection target, and calculating a priority of the predetermined selection target using the first index and the second index. The calculation device according to claim 1. The calculation unit calculates, for each selection target, a sum of a value obtained by adding a predetermined parameter to the first index and a value obtained by adding a predetermined parameter to the second index, and a value of the calculated sum The calculation device according to claim 9, wherein a priority for each of the selection targets is calculated based on the distribution of. The calculation unit learns a model for calculating the priority from the calculated sum value using the distribution of the calculated sum value as a feature, and calculates the priority for each selection target using the learned model The calculation device according to claim 10, wherein: The calculation unit calculates the priority of the predetermined selection target using the search query history including information related to the predetermined selection target among the search query histories in the web search as the input information. The calculation device according to claim 1, wherein The said calculation part calculates the priority of the said predetermined selection object based on the information regarding the said predetermined selection object posted on the web as said input information. Among Claims 1-12 characterized by the above-mentioned. The calculation device according to any one of the above. The calculation unit calculates the priority of the predetermined selection target using the purchase history of the predetermined selection target as the behavior information. The calculation device described. The acquisition unit acquires a set of the behavior information and a position where the behavior indicated by the behavior information is performed,
The calculation unit indicates, based on behavior information indicating a history of a user's behavior regarding selection of the predetermined selection target, and search query candidates including information related to the predetermined selection target. The calculation device according to any one of claims 1 to 13, wherein the priority of the predetermined selection target is calculated in consideration of weighting according to a position where the action is performed. A calculation method executed by a calculation device,
An acquisition process for acquiring behavior information indicating a history of user behavior related to selection of a selection target and a history of a search query input by the user;
Among the behavior information, behavior information indicating a history of user behavior related to selection of a predetermined selection target, and a search query history including information related to the predetermined selection target among the search query history. And a calculating step of calculating a priority when presenting the predetermined selection target. An acquisition procedure for acquiring behavior information indicating a history of a user's behavior related to selection of a selection target and a history of a search query input by the user;
Among the behavior information, behavior information indicating a history of user behavior related to selection of a predetermined selection target, and a search query history including information related to the predetermined selection target among the search query history. And a calculation procedure for causing a computer to execute a calculation procedure for calculating a priority when presenting the predetermined selection target.

Images