JP2019053520A - Providing apparatus, providing method, and providing program - Google Patents

Abstract

To make a proposal based on an arbitrary keyword. A providing apparatus according to the present application includes a first extracting unit, a second extracting unit, and a providing unit. A 1st extraction part extracts the some related query relevant to the input query input from the user based on the relevance degree of arbitrary queries. A 2nd extraction part extracts the same attribute query which is a query which has the attribute which is the same as or similar to an input query among the related queries extracted by the 1st extraction part. The providing unit provides information to the user based on the same attribute query extracted by the second extracting unit. [Selection] Figure 1

Description

  The present invention relates to a providing apparatus, a providing method, and a providing program.

  In recent years, with the development of communication networks, various services are provided via networks. In relation to this, various search techniques for searching for a service existing on a network or searching for a desired product or article in the service have been proposed.

  For example, a technique has been proposed that makes it possible to perform a quick search by performing a suggestion associated with a keyword appearing in a specific site and specializing in that site.

JP 2012-14208 A

  However, in the above-described conventional technology, it is not always possible to make a proposal based on an arbitrary keyword. Specifically, the above-described conventional technology enables a quick search by targeting keywords appearing in a specific site, for example, based on a wide variety of keywords with different fields and categories. It is not always possible to provide information such as advertisements, search suggestions, and recommendations.

  The present application has been made in view of the above, and an object thereof is to provide a providing device, a providing method, and a providing program capable of making a proposal based on an arbitrary keyword.

  The providing apparatus according to the present application extracts a plurality of related queries related to an input query input from a user based on a degree of association between arbitrary queries, and the first extraction unit extracts the related queries. Based on the same attribute query extracted by the 2nd extraction part which extracts the same attribute query which is a query which has the attribute which is the same as or similar to the said input query among related queries, and the said 2nd extraction part, with respect to the said user And a providing unit for providing information.

  According to one aspect of the embodiment, there is an effect that a proposal based on an arbitrary keyword can be performed.

FIG. 1 is a diagram illustrating an example of a providing process according to the embodiment. FIG. 2 is a diagram illustrating a configuration example of the providing system according to the embodiment. FIG. 3 is a diagram illustrating a configuration example of the providing apparatus according to the embodiment. FIG. 4 is a diagram illustrating an example of the association degree information storage unit according to the embodiment. FIG. 5 is a diagram illustrating an example of the attribute information storage unit according to the embodiment. FIG. 6 is a diagram illustrating an example of a recommendation storage unit according to the embodiment. FIG. 7 is a diagram for explaining processing of the first extraction unit according to the embodiment. FIG. 8 is a flowchart (1) illustrating a processing procedure performed by the providing apparatus according to the embodiment. FIG. 9 is a flowchart (2) illustrating a processing procedure performed by the providing apparatus according to the embodiment. FIG. 10 is a hardware configuration diagram illustrating an example of a computer that realizes the function of the providing apparatus.

  Hereinafter, embodiments for providing a providing apparatus, a providing method, and a providing program according to the present application (hereinafter referred to as “embodiments”) will be described in detail with reference to the drawings. In addition, the providing apparatus, the providing method, and the providing 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. Example of provision processing)
First, an example of a providing process according to the embodiment will be described with reference to FIG. FIG. 1 is a diagram illustrating an example of a providing process according to the embodiment. In FIG. 1, as an example of the providing process according to the embodiment, an example is a process in which the providing apparatus 100 according to the embodiment provides information according to a search query (hereinafter referred to as “query”) transmitted from the user U01. Will be described.

  A providing apparatus 100 illustrated in FIG. 1 is a server apparatus that provides information such as advertisements and recommendations based on a query transmitted from a user. In the embodiment, the providing device 100 also functions as a web server that provides a search service to the user.

  In the embodiment, a keyword used for a search in a search site or the like that provides a search service is referred to as a “query”. However, there may be a case where the keyword and the query do not completely match. For example, the user may input a sentence such as “AAA car” as a query (“AAA” is the name of an automobile manufacturer, for example). In this case, “AAA car” can be a query, and “AAA” and “car” that are words included in the query can be keywords. Providing apparatus 100 appropriately extracts keywords from the query, for example, by performing morphological analysis on the query. In the following description, “query” and “keyword” may be treated as synonymous. For example, when the query is composed of only one keyword or when one keyword is extracted from the result of morphological analysis of the query, the query and the keyword are treated as synonymous.

  The user group illustrated in FIG. 1 is a plurality of users who use a search site or the like, and is a plurality of users who transmit a query to the providing apparatus 100. The user U01 is an example of a user who transmits a query to the providing apparatus 100 (specifically, inputs a keyword at a search site or the like), and an advertisement or a recommendation according to the transmitted query. It is the user who becomes the providing destination to which the information is provided. Each user included in the user group and each of the users U01 use an information processing terminal (hereinafter referred to as “user terminal 10”) for using the search site and transmitting / receiving various types of information to / from the providing apparatus 100. Have). In the following description, the user terminal 10 and a user who uses the user terminal 10 may be identified with each other. For example, “the user U01 transmits a query” actually means that “the user terminal 10 used by the user U01 transmits a query”.

  A query transmitted when a user performs a search indicates the interest of the user. For this reason, by providing information such as advertisements and recommendations according to the query, it is possible to present the user with products and services similar to the object with which the user is interested. This is useful, for example, in enhancing marketing effects such as advertisement distribution and appeal to advertisements and recommendations. In other words, what kind of queries are similar to each other, and presenting the target corresponding to (similar to) the query transmitted by the user to the user is, for example, an advertisement distribution company, an advertisement, a recommendation, etc. This is desirable for providers who provide

  In addition, as a method of determining what queries are similar to each other, for example, a method of setting attributes (categories) for queries and classifying queries having similar attributes into the same class, etc. is there.

  However, the types of queries transmitted to search sites and the like are enormous, and it is not realistic to manually assign attributes. In addition, even in a method of assigning attributes to a query by a program or the like, there are cases where a plurality of meanings are included in one query, and it is difficult to automatically perform appropriate categorization.

  For example, when a keyword “A (“ A ”is an arbitrary word and indicates a movie title”) indicating a movie title is used as a query, it is expressed as “A that is a movie title”. It is difficult to determine whether it is used for the purpose of showing or the keyword “A” is used for other purposes. Specifically, when the user U01 inputs “A” as a query, when delivering a recommendation to the user U01 based on the query, it is effective to deliver a recommendation of any attribute (category). It is difficult to determine whether That is, when the providing apparatus 100 according to the embodiment distributes a recommendation regarding a shopping mall that handles products having various attributes and various fields, even if the recommendation device 100 simply provides a recommendation based on the keyword “A”, There is a case where the product or service indicated by the recommendation is not appropriate (for example, the product or service having the attribute desired by the user U01).

  Therefore, the providing apparatus 100 according to the embodiment executes a providing process described below. Specifically, the providing apparatus 100 calculates in advance the degree of association between two queries based on the number of times that the same user has input different queries among all the users to be aggregated. Thereafter, the providing apparatus 100 represents a query (hereinafter referred to as “related query”) related to an arbitrary query (hereinafter referred to as “input query”) input by the user based on the calculated degree of association between the queries. Multiple). Then, the providing apparatus 100 extracts a related query having an attribute that is the same as or similar to the input query (hereinafter referred to as “same attribute query”) from the extracted related queries. Furthermore, the providing apparatus 100 provides information such as advertisements and recommendations to the user who has input the input query based on the extracted identical attribute query.

  Accordingly, the providing apparatus 100 can select and provide a recommendation or the like based on a related query that matches the user's input query, even when the user inputs a keyword indicating any attribute as a query. Therefore, a recommendation according to the user's intention can be provided. For example, when the user inputs a query having “movie title” as an attribute, information related to other movie titles (that is, movies) can be provided to the user as recommended products or services for the user. That is, according to the providing apparatus 100, regardless of what keyword is input from the user, it is possible to provide information suitable for the input keyword (according to the user's intention). Hereinafter, an example of the providing process performed by the providing apparatus 100 will be described along the flow with reference to FIG.

  First, in the example illustrated in FIG. 1, the providing apparatus 100 calculates the degree of association between arbitrary queries. In the example shown in FIG. 1, it is assumed that each user in the user group inputs an arbitrary query at a search site or the like (step S11). Providing apparatus 100 acquires a query transmitted from each user (step S12). Then, the providing apparatus 100 calculates the degree of association between queries (step S13). In the following description, of the two queries whose degree of association is calculated, a query to be processed is referred to as a “first query”, and a query related to the first query is referred to as a “second query”.

  The providing apparatus 100 determines whether any query is based on at least the number of users who input any one of two different queries among the arbitrary queries and the number of users who input both of the two queries. The relevance of is calculated. As an example, the providing apparatus 100 associates the first query (“Query A” in Expression (1)) with the second query (“Query B” in Expression (1)) based on the following Expression (1). Calculate the degree.

  In the above equation (1), “Score (A, B)” indicates a numerical value of the degree of association between the query A and the query B. “Auser” indicates the number of users who have searched (input) query A. “Buser” indicates the number of users who searched the query B. “ALLuser” indicates the number of users (users using search) that transmitted a query during a predetermined counting period. “Auser∧Buser” indicates the number of users who have searched both query A and query B.

  The providing apparatus 100 calculates the relevance level of all queries searched from the user group using the above formula (1). For example, the providing apparatus 100 aggregates the query B searched by the user who has searched the query A that is the first query. Then, the providing apparatus 100 calculates the degree of relevance for all the second queries for the first query using the above formula (1). The providing apparatus 100 stores the calculated degree of association of the first query in the storage unit. The providing apparatus 100 may extract only a pair of the first query and the second query having a degree of association exceeding a predetermined threshold and store the extracted pair in the storage unit.

  As described above, the providing apparatus 100 according to the embodiment uses the query based on the quantitative information of the actual user search behavior, regardless of the qualitative information such as the meaning and nature of the query transmitted from the user. The relevance of is calculated. Specifically, since the providing apparatus 100 calculates the degree of relevance based on the user's search behavior log, the providing apparatus 100 can calculate the degree of relevance based on the actual interest of the user regardless of the meaning of the query itself. . In other words, the providing apparatus 100 does not simply calculate a high degree of association between queries having similar meanings of queries (keywords), but calculates a high degree of association between queries searched by users who take similar actions. Therefore, it is possible to calculate the degree of association in which the user's behavior is accurately reflected.

  Further, in the above equation (1), the providing apparatus 100 can adjust how long the query relevance is calculated in a period of time by changing the aggregation period. For example, if the providing period is set within a range of several years, the providing apparatus 100 may search for “first query and second query” when the same user searches for the first query and the second query during the several years. Since it counts as “the user who searched both of the queries”, it is possible to calculate the degree of relevance indicating the change of interest of the user in a relatively long range. Thereby, the providing apparatus 100 can calculate the degree of association that reflects the transition of the user's life stage (for example, the query that the user searches has changed from “pregnancy” to “birth”). On the other hand, if the providing apparatus 100 sets the aggregation period in a range of several days, the same user does not search both the first query and the second query during the several days. The number of users who have searched both of the second queries is not counted. For this reason, the providing apparatus 100 extracts a pair of related queries even when a relatively short period is set, instead of a pair of related queries when a relatively long period is set. It becomes easy. The aggregation period may be set as appropriate by, for example, an administrator of the providing apparatus 100.

  In the example of FIG. 1, the providing apparatus 100 stores the calculated degree of association in the database DB01. In the description of the embodiment, in order to clearly indicate an attribute for each keyword (query), one keyword (query) is described with a pair of a concept code indicating the attribute and the keyword. For example, when the query “A” is a movie title, the query is described as “movie title A”.

  In the example shown in FIG. 1, the database DB01 has a case where the first query is “movie title A” and the second query is “director name D”, “movie introduction site E”, and “movie title F”, respectively. The relevance is memorized. Specifically, as an example of information stored in the database DB01, the relevance level of a pair whose first query is “movie title A” and whose second query is “director name D” is “4.21”. It shows that there is. As another example of information stored in the database DB01, the degree of association of a pair in which the first query is “movie title B” and the second query is “movie title G” is “4.15”. It is shown that. As described above, since the degree of association between queries is proportional to the number of users who input queries, the same second query appears repeatedly in the second query for different first queries. There may be cases. For example, as shown in FIG. 1, “movie title F” may appear as a second query common to both the first query “movie title A” and the first query “movie title B”.

  The providing apparatus 100 retains the degree of relevance between queries as shown in the database DB01 for all queries input from the user group.

  After that, the providing apparatus 100 receives an input of an arbitrary query from the user U01 (Step S14). For example, the user U01 uses the user terminal 10 to input “movie title A” and “movie title B” as input queries in a search window displayed on the top page of a shopping mall provided by the providing device 100. input.

  In this case, the providing apparatus 100 refers to the database DB01 and extracts a plurality of related queries related to the input query (step S15). Specifically, providing apparatus 100 extracts a plurality of related queries for “movie title A” and “movie title B”, which are input queries of user U01.

  In the example of FIG. 1, the providing apparatus 100 sets “director name D”, “movie introduction site E”, and “movie title F” as related queries in the descending order of the degree of relevance as related queries for “movie title A”. Extract. Further, the providing apparatus 100 extracts “movie title G”, “actor H”, and “movie title F” as related queries in descending order of the degree of relevance as related queries for “movie title B”.

  Furthermore, the providing apparatus 100 further extracts a related query (same attribute query) having an attribute that is the same as or similar to the input query from among the plurality of related queries extracted in step S15 (step S16).

  Although the details of the extraction method in step S16 will be described later, as an example, the providing apparatus 100 uses the same attribute query of the input query as a related query that is associated with a keyword that is the same as or similar to the keyword associated with the input query. Extract as More specifically, the providing apparatus 100 uses the related query extracted in step S15 as a search suggestion using a keyword that is the same as or similar to the keyword used as the search suggestion for the input query in a predetermined search engine. Are extracted as the same attribute query of the input query.

  For example, the providing apparatus 100 refers to the information that “Synopsis” appears as a higher-level suggest word when “movie title A” is input in a predetermined search engine. At this time, the providing apparatus 100 sets a related query in which “Synopsis” appears as a suggestion word to “Movie Title A” when an arbitrary “Related Query for“ Movie Title A ”” is input in a predetermined search engine. Are extracted as candidates for the same attribute query. This is because if the query has the same or similar attribute as the input query, it is presumed that the probability that the search engine has a common suggestion word is high. Specifically, if the query has an attribute of “movie title”, it is estimated that “summary” appears in the suggestion word, but if the attribute is “director name” or “actor” Therefore, it is estimated that there is a low possibility that the “summary” appears in the suggestion word. Based on the estimation, the providing apparatus 100 determines whether or not the attributes of the input query and the related query match based on the possibility of appearance of the same suggestion word.

  As will be described later, the providing apparatus 100 may perform a process of extracting the same attribute query of the input query (in other words, filtering only the same attribute query from the related query) by a method different from the above. Further, as illustrated in FIG. 1, the providing apparatus 100 may extract the same attribute query having attributes common to the plurality of input queries when a plurality of input queries are received from the user U01. .

  In the example of FIG. 1, the providing apparatus 100 uses “movie title F” as a query having an attribute (in this example, “movie title”) common to “movie title A” and “movie title B” received from the user U01. And “movie title G” are extracted.

  Subsequently, the providing apparatus 100 refers to the database DB02 in which the same attribute query and the recommendation information are associated with each other, and specifies the recommendation information associated with the related query extracted in step S16 (that is, the same attribute query). (Step S17). Specifically, the providing apparatus 100 identifies “image data 50” that is recommendation information of “movie title F” that is the extracted identical attribute query. Further, the providing apparatus 100 identifies “image data 52” that is recommendation information of “movie title G” that is the extracted identical attribute query. For example, the image data 50 is image data extracted from a product introduction page of a medium in which a movie corresponding to “movie title F” is stored, and the image data 52 stores a movie corresponding to “movie title G”. It is the image data extracted from the product introduction page of the selected media.

  Then, the providing apparatus 100 transmits the information specified in step S17 to the user U01. That is, the providing apparatus 100 provides the user U01 with the image data 50 and the image data 52 that are recommendation information (step S18). As a result, the user U01 can receive recommendation information related to “movie title F” and “movie title G” having attributes common to the movie title A and movie title B input by the user U01.

  As described above with reference to FIG. 1, the providing apparatus 100 according to the embodiment extracts a plurality of related queries related to the input query input from the user U01 based on the degree of association between arbitrary queries. Then, the providing apparatus 100 extracts the same attribute query that is a query having the same or similar attribute as the input query from the extracted related queries. Furthermore, the providing apparatus 100 provides information to the user U01 based on the extracted identical attribute query.

  That is, according to the providing apparatus 100, when an arbitrary query is input from the user U01, a query (which may be read as a keyword) that is the same as or similar to the query is extracted, and the user is based on the extracted query. Specify advertisements and recommendation information to be provided to U01. Accordingly, the providing apparatus 100 can identify recommendation information according to the intention of the user U01 and the interest of the user U01 and provide it to the user regardless of what kind of query is input from the user U01. it can. Thus, the providing apparatus 100 can make a proposal based on an arbitrary keyword. Hereinafter, the providing apparatus 100 that performs the processing as described above and the providing system 1 including the providing apparatus 100 will be described in detail.

[2. System configuration
Next, the configuration of the providing system 1 according to the embodiment will be described with reference to FIG. FIG. 2 is a diagram illustrating a configuration example of the providing system 1 according to the embodiment. As shown in FIG. 2, the providing system 1 includes a user terminal 10 and a providing device 100. The user terminal 10 and the providing apparatus 100 are connected to be communicable by wire or wireless via a network N (for example, the Internet) that is a communication network. Note that the number of devices included in the providing system 1 illustrated in FIG. 2 is not limited to that illustrated. For example, the providing system 1 may include a plurality of user terminals 10 and the like.

  The user terminal 10 is an information processing device used by a user. For example, the user terminal 10 is an information processing apparatus such as a desktop PC, a notebook PC, a mobile phone such as a smartphone, a tablet terminal, a PDA (Personal Digital Assistant), or a wearable device. For example, the user terminal 10 accesses a site or the like that provides a search service according to a user operation. Then, the user terminal 10 transmits the query input by the user to a server (providing device 100 in the embodiment) that provides the search service.

  As described above, the providing apparatus 100 extracts a plurality of related queries related to the input query, extracts the same attribute query that is a query having the same or similar attribute as the input query, and extracts the extracted same attribute query. This is a server device that provides information to the user based on the information.

[3. Configuration of the providing device]
Next, the configuration of the providing apparatus 100 according to the embodiment will be described with reference to FIG. FIG. 3 is a diagram illustrating a configuration example of the providing apparatus 100 according to the embodiment. As illustrated in FIG. 3, the providing apparatus 100 includes a communication unit 110, a storage unit 120, and a control unit 130. The providing apparatus 100 includes an input unit (for example, a keyboard and a mouse) that receives various operations from an administrator who uses the providing apparatus 100, and an output unit (for example, a liquid crystal display) for outputting various types of information. You may have.

(About the communication unit 110)
The communication unit 110 is realized by, for example, a NIC (Network Interface Card). The communication unit 110 is connected to the network N by wire or wireless, and transmits / receives information to / from the user terminal 10 via the network N.

(About the storage unit 120)
The storage unit 120 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. The storage unit 120 according to the embodiment includes a relevance information storage unit 121, an attribute information storage unit 122, and a recommendation storage unit 123. Hereinafter, each storage unit will be described in order.

(Regarding Relevance Information Storage Unit 121)
The relevance information storage unit 121 stores information related to the relevance between queries. The relevance information storage unit 121 corresponds to the database DB01 shown in FIG. Here, FIG. 4 shows an example of the relevance information storage unit 121 according to the embodiment. FIG. 4 is a diagram illustrating an example of the association degree information storage unit 121 according to the embodiment. As shown in FIG. 4, the relevance information storage unit 121 includes items such as “total period”, “total number of search users”, “first query”, “second query”, and “relevance”.

  “Aggregation period” indicates a period during which information related to user search behavior is aggregated. “Total number of search users” indicates information related to the total number of search users in the counting period. In FIG. 4, the information stored in the “total number of search users” is conceptually shown as “A01”, but in reality, the item “total number of search users” A specific numerical value of the total number of unique users using the search service to be processed is stored. The item may store information such as the number of times a query has been transmitted to the search service (in other words, a search has been performed) during the counting period, and the number of searches for each query.

  The “first query” indicates the first query. The “second query” is a query searched by the user who has searched for the first query, and indicates a query different from the first query. “Relevance” indicates the relevance between the first query and the second query. The degree of association is calculated for each pair of queries by the above formula (1), for example.

  That is, in FIG. 4, as an example of information held in the relevance information storage unit 121, the aggregation period is “July 1, 2016 to June 30, 2017”, and the total number of search users is “A01”. This indicates that the relevance of the pair whose first query is “movie title A” and whose second query is “director name D” is “4.21”.

(About the attribute information storage unit 122)
The attribute information storage unit 122 stores attribute information of the query. Here, FIG. 5 illustrates an example of the attribute information storage unit 122 according to the embodiment. FIG. 5 is a diagram illustrating an example of the attribute information storage unit 122 according to the embodiment. As illustrated in FIG. 5, the attribute information storage unit 122 includes items such as “query” and “attribute”.

  “Query” indicates a query associated with an attribute. In the example of FIG. 5, a query can be read as a keyword. “Attribute” indicates an attribute associated with the query.

  In the embodiment, the query attribute is information indicating the characteristics and classification of each query, and indicates, for example, a concept that is given artificially or determined by an extraction process described later. The example shown in FIG. 5 is an example in which a query to which an attribute is assigned is held as a database, and such an example may be read as dictionary data that stores the relationship between the query and the attribute.

  In the example shown in FIG. 5, the query is classified into attributes such as “movie”, “person”, and “information content”, but the attributes are not limited to the example shown here. That is, as the attribute, any information may be used as long as the information can classify (cluster) the query for each feature of the query. Further, the attribute may use information held in an external knowledge database (for example, a knowledge base), or may be uniquely set by an administrator of the providing apparatus 100 or the like.

  Further, the attribute is not limited to the example shown in FIG. 5, and concepts that are classified in more detail may be used. For example, the providing apparatus 100 may associate the attribute of the movie title A with “movie title” instead of “movie”, and associate the attribute of the director name D with “movie director” instead of “person”. May be.

  Whether or not the attributes are similar may be automatically determined based on, for example, the similarity between the queries, or the providing apparatus 100 may be based on the similarity of the concept assigned as the attribute. It may be determined by the administrator or the like. Specifically, the providing apparatus 100 determines whether or not a predetermined number or more users out of the total number of users are based on the number of users searching for a query having the attribute “movie” and a query having the attribute “music”. It may be determined that the attributes are similar). Alternatively, the providing apparatus 100 may hold definition data indicating that the attribute “movie” and the attribute “music” are similar in the attribute information storage unit 122 and perform determination based on the definition data.

  That is, in FIG. 5, as an example of information held by the attribute information storage unit 122, the attribute associated with the query “movie title A” is “movie”. Further, it is indicated that the attribute associated with the query “director name D” is “person”.

(Recommendation storage unit 123)
The recommendation storage unit 123 stores recommendation information provided to the user. The recommendation storage unit 123 corresponds to the database DB02 shown in FIG. Here, FIG. 6 illustrates an example of the recommendation storage unit 123 according to the embodiment. FIG. 6 is a diagram illustrating an example of the recommendation storage unit 123 according to the embodiment. As illustrated in FIG. 6, the recommendation storage unit 123 includes items such as “query” and “recommendation information”.

  “Query” indicates a query associated with recommendation information. In the example of FIG. 6, a query can be read as a keyword. “Recommendation information” indicates recommendation information associated with a query. The recommendation information is data serving as a material displayed as recommendation information in a service accessed by the user, and is, for example, image data or link information embedded in the image data. Note that the recommendation information may be generated in real time based on the query without being stored in the recommendation storage unit 123. For example, with respect to the query “movie title A”, the providing apparatus 100 recommends information based on image data posted on a product or service introduction page of “movie title A” on a predetermined shopping site and link information on the page. May be generated in real time.

  That is, FIG. 6 shows that the recommendation information associated with the query “movie title A” is “image data K01” and “link L01” as an example of information stored in the recommendation storage unit 123. Yes.

(About the control unit 130)
The control unit 130 is a controller, for example, various programs (an example of a provided program) stored in a storage device inside the providing apparatus 100 by a CPU (Central Processing Unit), an MPU (Micro Processing Unit), or the like. This is realized by executing the RAM as a work area. The control unit 130 is a controller, and is realized by an integrated circuit such as an application specific integrated circuit (ASIC) or a field programmable gate array (FPGA).

  As illustrated in FIG. 3, the control unit 130 according to the embodiment includes an acquisition unit 131, a calculation unit 132, a reception unit 133, a first extraction unit 134, a second extraction unit 135, and a provision unit 136. And realizes or executes the functions and operations of information processing described below. Note that the internal configuration of the control unit 130 is not limited to the configuration illustrated in FIG. 3, and may be another configuration as long as the information processing described below is performed. Further, the connection relationship between the processing units included in the control unit 130 is not limited to the connection relationship illustrated in FIG. 3, and may be another connection relationship.

(About the acquisition unit 131)
The acquisition unit 131 acquires various types of information. For example, the acquisition unit 131 acquires a query transmitted from the user. Specifically, the acquisition unit 131 acquires a keyword that is arbitrarily input from a plurality of users such as the user group illustrated in FIG. 1 and is used as a query for search processing at a search site or the like. Note that the query may include a plurality of keywords.

  In addition, the acquisition unit 131 acquires the number of users who input one of two different queries among arbitrary queries and the number of users who input both of the two queries in a predetermined aggregation period. To do. For example, the acquisition unit 131 is based on information for identifying a user who has searched for the first query (a user ID in a service, a terminal ID that is identification information unique to a terminal, a cookie of a browser software, etc.) It can be determined whether the user has searched for the second query.

  Further, the acquisition unit 131 acquires the total number of searches and the total number of unique users in a predetermined counting period. Note that the acquisition unit 131 may acquire the information from one search site, or may add the information acquired from a plurality of search sites. In this case, the acquisition unit 131 may acquire information on the search behavior of the user from a predetermined external server that provides the search site to the user. Moreover, the acquisition part 131 may acquire the query transmitted from a user not only in a search site but in a predetermined service site (for example, a shopping site or an auction site).

  In addition, the acquisition unit 131 may acquire attribute information set in the query (keyword). For example, the acquisition unit 131 acquires attribute information set in the query based on an input by the administrator of the providing apparatus 100 or the like. Alternatively, the acquisition unit 131 may acquire attribute information associated with the query from an external knowledge database or the like by a predetermined crawl process (crawling).

  Moreover, the acquisition part 131 may acquire the recommendation information etc. which are provided to a user from the provider who provides an advertisement and a recommendation, an advertisement delivery company, etc. For example, the acquisition unit 131 acquires image data and moving image data related to a query, text data describing advertisements and product introductions, link information, and the like as recommendation information.

  The acquisition unit 131 stores the acquired information in the storage unit 120. For example, the acquisition unit 131 describes the acquired attribute information in the attribute information storage unit 122 or describes the acquired recommendation information in the recommendation storage unit 123. Further, the acquisition unit 131 may appropriately acquire information required by each processing unit described later from the storage unit 120.

(About the calculation unit 132)
The calculation unit 132 calculates the degree of association between the queries acquired by the acquisition unit 131. Specifically, the calculation unit 132 is based at least on the number of users who input any one of two different queries among arbitrary queries and the number of users who input both of the two queries. The relevance between arbitrary queries is calculated. Note that when the query includes a plurality of keywords, the calculation unit 132 may count the number of users who input the keywords for each keyword.

  In addition, the calculation unit 132 may calculate the degree of association based on the number of users who input both two different queries within a predetermined period. For example, when the aggregation period is set, the calculation unit 132 calculates the degree of association based on the number of users who input both two different queries during the aggregation period. In this way, the calculation unit 132 calculates the relevance that captures a relatively long period (such as one year or more) by making the period for collecting information that two queries are searched variable, or It is possible to control whether or not to calculate the degree of relevance that captures a relatively short period.

  Note that the calculation unit 132 may classify users who input a query and calculate the degree of association for each group that has been classified. For example, the calculation unit 132 classifies users into a predetermined group based on at least one of a predetermined query input history, a predetermined service usage history, and a predetermined attribute, and relates to each classified group. The degree may be calculated. For example, the calculation unit 132 may calculate the degree of association between queries in a user group having a history of searching for queries related to life stages such as “pregnancy” and “birth” a predetermined number of times or more. Alternatively, the calculation unit 132 may calculate the degree of association between queries in a user group having a history of using a shopping service or an auction service for a predetermined number of times or a user group having a common age group or gender. Thereby, the calculation part 132 can calculate the relevance degree of a query not only based on the search behavior of a large number of unspecified users but also based on the search behavior of users in a group having a specific interest.

  The calculation unit 132 calculates the degree of association between the first query and the second query, which are two different queries, by substituting a numerical value corresponding to the formula shown by the formula (1). The calculation unit 132 associates the pair of the first query and the second query and the calculated degree of association with each other and stores them in the degree-of-association information storage unit 121.

(About the reception unit 133)
The accepting unit 133 accepts various requests. For example, the reception unit 133 receives an arbitrary keyword from a predetermined user. Specifically, the accepting unit 133 accepts a search request that uses an arbitrary keyword transmitted by a user of the search service as a query in the search service. The receiving unit 133 may receive a query including a plurality of keywords as well as a single keyword. The accepting unit 133 sends the accepted query to the first extracting unit 134.

(About the first extraction unit 134)
The 1st extraction part 134 extracts the some related query relevant to the input query input from the user based on the relevance degree of arbitrary queries.

  Specifically, the first extraction unit 134 refers to the relevance information storage unit 121 in which the relevance calculated by the calculation unit 132 is held with the input query as the first query. And the 1st extraction part 134 is a 2nd query corresponding to an input query, Comprising: A predetermined number is extracted as a related query of an input query in an order from the 2nd query with high relevance with an input query. In addition, the number extracted as a related query can be set arbitrarily.

  In addition, when a plurality of input queries are input from the user, the first extraction unit 134 extracts a plurality of related queries related to each input query, and each of the plurality of input queries and the plurality of input queries. The sum of the degree of relevance with the related query is calculated, and a predetermined number of related queries are extracted based on the calculated value. This point will be described with reference to FIG. FIG. 7 is a diagram for explaining the processing of the first extraction unit 134 according to the embodiment. In FIG. 7, the database DB 11 that shows information used for processing by the first extraction unit 134 as a table will be described as an example. The database DB11 is generated based on information stored in the relevance information storage unit 121, for example.

  FIG. 7 shows information related to related query extraction processing when the reception unit 133 receives three queries, “movie title X”, “movie title Y”, and “movie title Z” as input queries. . In the database DB11, the degree of association between “movie title X” and the second query “movie title M” is “5.73”, and “movie title Y” and the second query “movie title M”. Is “5.03”, and the degree of association between “movie title Z” and the second query “movie title M” is “4.26”. In addition, the database DB 11 indicates that the sum of the degrees of relevance between the three queries “movie title X”, “movie title Y”, and “movie title Z” and “movie title M” is “15.02”. Show.

  As illustrated in FIG. 7, when a plurality of input queries are received, the first extraction unit 134 calculates the sum of the degrees of association between each input query and the second query. Then, the first extraction unit 134 extracts, as related queries, a predetermined number of second queries with the highest relevance level. In the example of FIG. 7, the first extraction unit 134 includes “movie title M” and “actor N” as related queries of three queries “movie title X”, “movie title Y”, and “movie title Z”. "Director name O", "Movie title P", "Actor Q", "Movie title R", "Movie title S", "Movie title T", "Shooting location U", etc. Indicates that it has been extracted.

  As described above, the degree of relevance by the calculation unit 132 is calculated based on the number of users who search for a certain two queries together. For this reason, the related query shown in FIG. 7 is, in other words, a target that a user who is interested in any of the three queries “movie title X”, “movie title Y”, and “movie title Z” is interested in. It can be said that

  The first extraction unit 134 may extract a predetermined number of second queries as related queries in descending order of relevance, or all second queries having a relevance exceeding a predetermined threshold as related queries. It may be extracted. The predetermined threshold value may be arbitrarily set by the administrator of the providing apparatus 100, for example, or may be calculated by a statistical method (for example, an average value of all calculated degrees of association is used as the predetermined threshold value). Etc.) The first extraction unit 134 sends information related to the extracted related query to the second extraction unit 135.

(About the second extraction unit 135)
The second extraction unit 135 extracts the same attribute query that is a query having an attribute that is the same as or similar to the input query from the related queries extracted by the first extraction unit 134.

  The second extraction unit 135 can extract the same attribute query of the input query by various methods. For example, the second extraction unit 135 may extract the same attribute query based on dictionary data in which an attribute of an arbitrary query is defined in advance. In this case, the second extraction unit 135 identifies the attribute of the input query based on dictionary data (for example, information stored in the attribute information storage unit 122). And the 2nd extraction part 135 extracts the query which is the same as or similar to the attribute of an input query among related queries as an identical attribute query.

  In addition, the second extraction unit 135 selects a related query that is associated with a keyword that is the same as or similar to the keyword that is associated with the input query among the related queries extracted by the first extraction unit 134. It may be extracted as an attribute query. In this case, the second extraction unit 135 does not need to specify the attribute of the input query. That is, the second extraction unit 135 can extract the same attribute query based on whether the same or similar keyword is associated with the keyword associated with the input query regardless of the attribute of the input query. .

  Specifically, the second extraction unit 135 uses, as a search suggestion, a keyword that is the same as or similar to a keyword used as a search suggestion for an input query in a predetermined search engine among the related queries extracted by the first extraction unit 134. The related query to be used is extracted as the same attribute query of the input query. Here, the search suggestion means a word (so-called suggest word) that is proposed to the user as a word that can be easily searched when an arbitrary query is input to the search engine.

  For example, the second extraction unit 135 may use “summary”, “release date”, “theatre”, “mobilization” as a suggested word when “(movie title) A” is input in a predetermined search engine. "Is included. Then, the second extraction unit 135, as a suggestion word when “F” which is one of the related queries of “(movie title) A” is input in a predetermined search engine, When “release date”, “theatre”, “mobilization”, etc. are included, it is determined that “A” and “F” are queries having the same attribute (in this example, attribute “movie”). . Then, the second extraction unit 135 extracts “(movie title) F” from the related queries as the same attribute query of the input query “(movie title) A”.

  The second extraction unit 135 is not limited to the search suggestion as a keyword associated with the query, and various keywords can be used. For example, the second extraction unit 135 may extract the same attribute query using a word (so-called co-occurrence word) that appears with the query in a predetermined document.

  In addition, the second extraction unit 135 is based on the similarity between each content reached from the search result based on each related query extracted by the first extraction unit 134 and the content reached from the search result based on the input query. You may extract the same attribute query of the said input query. The content reached from the search result is, for example, a web page displayed as the search result when a related query is input to the search engine.

  Specifically, the second extraction unit 135 is similar to the domain of the page reached from the search result based on each related query extracted by the first extraction unit 134 and the domain of the page reached from the search result based on the input query. Based on the degree, the same attribute query of the input query is extracted. For example, when “(movie title) A” is inputted in a predetermined search engine, an official page of “A” that is a movie, an information site introducing various movies, an official movie theater It is estimated that a site or the like is displayed as a search result. At this time, even when “(movie title) F”, which is a related query of “(movie title) A”, is input, the second extraction unit 135 may be an information site introducing various movies or a predetermined movie. Reference is made to searching for content common to “(movie title) A”, such as the official website of the theater. Then, the second extraction unit 135, when the similarity between the domain of the page reached from the search result based on the related query and the domain of the page reached from the search result based on the input query exceeds a predetermined threshold (for example, It is determined that the related query and the input query have the same or similar attributes (such as when a predetermined number or more of matching domains are included in the top 100).

  In addition, the second extraction unit 135 uses the learning model that has learned the relationship with the query having the same or similar attribute as the input query, and among the related queries extracted by the first extraction unit 134, The same attribute query may be extracted. As an example, the second extraction unit 135 generates, for all queries acquired by the acquisition unit 131, a feature vector having a high degree of relevance as a dimension and a degree of relevance as the number of dimensions. In addition, the second extraction unit 135 acquires several pairs of queries having the same attribute as correct answer data. Then, the second extraction unit 135 calculates the similarity between the feature vectors of the two queries (for example, calculates the cosine similarity of the two queries) and how the similarity is related to the correct data. It learns whether it has sex and generates a learning model.

  According to such a learning model, when two queries are input, the similarity between the feature vectors of the two queries is calculated, and the attribute that the two queries are the same or similar is calculated based on the calculated similarity. The likelihood of having or not having is output. The second extraction unit 135 determines whether the two queries are queries having the same or similar attributes based on the output likelihood. Specifically, when a pair of an input query and a related query is input to the learning model, the second extraction unit 135 outputs a likelihood that the input query and the related query are the same or similar attributes. Then, based on the output likelihood, the second extraction unit 135 determines whether the input query and the related query have the same or similar attributes. When it is determined that the related query has the same or similar attribute as the input query, the second extraction unit 135 extracts the related query as the same attribute query. Note that the second extraction unit 135 may use various known learning methods in generating the learning model. For example, the second extraction unit 135 may generate a learning model using deep learning, a support vector machine, a predetermined regression equation, or the like.

  Note that the second extraction unit 135 extracts a predetermined number of related queries based on the sum of the degrees of association between each of the plurality of input queries and the plurality of related queries by the first extraction unit 134. The same attribute query having the same or similar attribute as at least one input query of the plurality of input queries may be extracted from the predetermined number of related queries extracted by the first extraction unit 134. In the example of FIG. 7, the second extraction unit 135 includes “movie”, which is an attribute common to the three queries of the input query “movie title X”, “movie title Y”, and “movie title Z” among the related queries. “Movie title M”, “movie title P”, “movie title R”, “movie title S”, and “movie title T” are further extracted.

  Note that the second extraction unit 135 does not necessarily need to extract a query having an attribute common to all of the plurality of input queries as the same attribute query. That is, the user does not always input a query having the same attribute as the three queries “movie title X”, “movie title Y”, and “movie title Z” in the search service. For example, the user may input a query having attributes that are not the same as or similar to each other, such as three queries “movie title X”, “summary”, and “release date”. In this case, the second extraction unit 135 extracts the same attribute query having the same or similar attribute as at least one input query.

  Specifically, the second extraction unit 135 may extract the same attribute query of the input query first input from the user from the predetermined number of related queries extracted by the first extraction unit 134. Good. This is because, in general, the query that is input first often includes an object that most reflects the interest of the user.

  The second extraction unit 135 extracts the same attribute query of the input query by combining any of the above-described methods or any of the above-described methods. Then, the second extraction unit 135 sends the extracted identical attribute query to the providing unit 136.

(About the providing unit 136)
The providing unit 136 provides information to the user based on the same attribute query extracted by the second extracting unit 135.

  Specifically, the providing unit 136 refers to the recommendation storage unit 123 and identifies recommendation information associated with a query corresponding to the same attribute query. Then, the providing unit 136 provides a user with a recommendation related to the same attribute query extracted by the second extracting unit 135. Note that the information provided by the providing unit 136 is not limited to a recommendation but may be an advertisement or the like.

  The providing unit 136 may optimize the recommended information to be provided based on a known advertisement distribution technique or the like. For example, the providing unit 136 acquires index values such as CTR (Click Through Rate) and CVR (Conversion Rate) of the provided recommendation information, and based on the acquired information, the provision priority of recommendation information with a low appeal effect May be set low, or feedback regarding the correctness of the attribute determination processing by the second extraction unit 135 may be performed.

[4. Processing procedure)
Next, a processing procedure by the providing apparatus 100 according to the embodiment will be described with reference to FIGS. 8 and 9. First, the procedure of the relevance calculation process according to the embodiment will be described with reference to FIG. FIG. 8 is a flowchart (1) illustrating a processing procedure performed by the providing apparatus 100 according to the embodiment.

  As illustrated in FIG. 8, the providing apparatus 100 determines whether a query transmitted from the user has been acquired (step S101). When the query is not acquired (step S101; No), the providing apparatus 100 waits until the query is acquired.

  On the other hand, when a query is acquired (step S101; Yes), the providing apparatus 100 selects an arbitrary first query to be processed from the acquired queries (step S102). Furthermore, the providing apparatus 100 aggregates the second query searched by the user who searched for the first query (step S103).

  Then, the providing apparatus 100 calculates the degree of association between the first query and the second query using, for example, the above equation (1) (step S104). After that, the providing apparatus 100 determines whether or not the relevance levels of all search queries have been calculated (step S105). When the relevance levels of all the queries are not calculated (step S105; No), the providing apparatus 100 repeats the processing from step S102 to step S104.

  On the other hand, when the relevance level of all the queries is calculated (step S105; Yes), the providing apparatus 100 stores the calculated relevance level in the storage unit 120 (step S106).

  Next, the procedure of the providing process according to the embodiment will be described with reference to FIG. FIG. 9 is a flowchart (2) illustrating a processing procedure performed by the providing apparatus 100 according to the embodiment.

  As illustrated in FIG. 9, the providing apparatus 100 determines whether an input query has been received from the user (step S201). When the input query has not been received (step S201; No), the providing apparatus 100 waits until it is received.

  On the other hand, when an input query is received (step S201; Yes), the providing apparatus 100 extracts a plurality of related queries related to the input query based on the degree of association (step S202).

  Subsequently, the providing apparatus 100 further extracts the same attribute query having the same or similar attribute as the input query from the plurality of related queries extracted in step S202 (step S203).

  Then, the providing device 100 provides a recommendation to the user who is the transmission source of the input query based on the same attribute query extracted in step S203 (step S204).

[5. (Modification)
The providing system 1 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 each device included in the providing system 1 will be described.

[5-1. Morphological analysis for queries)
In the above-described embodiment, the example in which the providing apparatus 100 calculates the degree of association between queries based on the number of users who input both the first query and the second query has been described. Here, as described above, a query may be composed of not only one keyword but also a plurality of keywords and sentences. For this reason, the providing apparatus 100 may perform a morphological analysis on the query transmitted from the user, and handle a keyword included in the result of the morphological analysis as the first query or the second query. In this case, the providing apparatus 100 may extract nouns and proper nouns included in the query using a known description, and use only the extracted keywords for processing.

[5-2. Search behavior)
Providing apparatus 100 does not necessarily require that the search action is performed in the same period as the aggregation period for the period for determining that the user has searched both the first query and the second query. That is, the providing apparatus 100 may set a period for determining that the user has searched both the first query and the second query, and a period for counting the number of users who have searched the query. For example, when the first query and the second query are transmitted from the same user within 24 hours, the providing apparatus 100 treats the user as a user who has searched both the first query and the second query. Good. In addition, when the first query and the second query are transmitted in the same session for the same user (for example, a series of actions until access to a predetermined search site and access is interrupted), the providing apparatus 100 May be treated as a user who has searched both the first query and the second query. As described above, the providing apparatus 100 may handle the user's search behavior flexibly and perform various information processing.

[5-3. Search site)
In the above-described embodiment, an example in which a search site is provided by the providing device 100 has been described. However, the search site may be provided by a predetermined external server (for example, a web server that provides a search service). In this case, the providing apparatus 100 may acquire a query, user information, and the like transmitted by the user to the search site via an external server.

[6. Others]
In addition, among the processes described in the above embodiment, all or part of the processes described as being automatically performed can be performed manually, or the processes described as being performed manually can be performed. All or a part can be automatically performed by a known method. In addition, the processing procedures, specific names, and information including various data and parameters shown in the document and drawings can be arbitrarily changed unless otherwise specified.

  Further, each component of each illustrated apparatus is functionally conceptual, and does not necessarily need to be physically configured as illustrated. In other words, the specific form of distribution / integration of each device is not limited to that shown in the figure, and all or a part thereof may be functionally or physically distributed or arbitrarily distributed in arbitrary units according to various loads or usage conditions. Can be integrated and configured.

  For example, the relevance information storage unit 121, the attribute information storage unit 122, and the recommendation storage unit 123 illustrated in FIG. 3 may be stored in an external storage server or the like without being provided by the providing apparatus 100. In this case, the providing apparatus 100 obtains relevance information and feature information by accessing the storage server.

  In addition, for example, the providing device 100 described above has the same attributes as those of the front-end server that mainly executes exchanges with external devices such as acquiring a query from the user terminal 10 or transmitting provided information to the user. It may be distributed to the back-end server side that executes the process of specifying the provided information based on the query.

[7. Hardware configuration)
Further, the providing apparatus 100 and the user terminal 10 according to the above-described embodiments are realized by a computer 1000 having a configuration as shown in FIG. 10, for example. Hereinafter, the providing apparatus 100 will be described as an example. FIG. 10 is a hardware configuration diagram illustrating an example of a computer 1000 that implements the functions of the providing apparatus 100. The computer 1000 includes a CPU 1100, a RAM 1200, a ROM (Read Only Memory) 1300, an HDD (Hard Disk Drive) 1400, a communication interface (I / F) 1500, an input / output interface (I / F) 1600, and a 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 shown in FIG. 2) and sends the data to the CPU 1100, and other data generated by the CPU 1100 via the communication network 500 To the device.

  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 providing apparatus 100, the CPU 1100 of the computer 1000 implements the function of the control unit 130 by executing a program loaded on the RAM 1200. In addition, each data in the storage unit 120 is stored in the HDD 1400. The CPU 1100 of the computer 1000 reads these programs from the recording medium 1800 and executes them, but as another example, these programs may be acquired from other devices via the communication network 500.

[8. effect〕
As described above, the providing apparatus 100 according to the embodiment includes the first extracting unit 134, the second extracting unit 135, and the providing unit 136. The 1st extraction part 134 extracts the some related query relevant to the input query input from the user based on the relevance degree of arbitrary queries. The second extraction unit 135 extracts the same attribute query that is a query having an attribute that is the same as or similar to the input query from the related queries extracted by the first extraction unit 134. The providing unit 136 provides information to the user based on the same attribute query extracted by the second extracting unit 135.

  As described above, when an arbitrary query is input from the user, the providing apparatus 100 according to the embodiment extracts a query that is the same as or similar to the query, and provides information to the user based on the extracted query. I do. Accordingly, the providing apparatus 100 can provide information in which the user's intention and the user's interests are appropriately reflected regardless of what kind of query is input from the user. Thus, the providing apparatus 100 can make a proposal based on an arbitrary keyword.

  Further, the second extraction unit 135 extracts the same attribute query based on dictionary data in which an attribute of an arbitrary query is defined in advance.

  As described above, the providing apparatus 100 according to the embodiment can perform extraction processing with high accuracy by determining an attribute based on dictionary data in which the definition of the attribute is held in advance.

  In addition, the second extraction unit 135 extracts, from the related queries extracted by the first extraction unit 134, related queries that are associated with keywords that are the same as or similar to the keywords associated with the input query. Extract as the same attribute query.

  As described above, the providing apparatus 100 according to the embodiment may determine whether or not the input query has the same attribute based on the commonality of related keywords, and extract the same attribute query. As a result, the providing apparatus 100 can extract the same attribute query with high accuracy without holding definition data such as dictionary data.

  In addition, the second extraction unit 135 uses, as a search suggestion, a keyword that is the same as or similar to a keyword used as a search suggestion for an input query in a predetermined search engine among the related queries extracted by the first extraction unit 134. The query is extracted as the same attribute query of the input query.

  As described above, the providing apparatus 100 according to the embodiment can extract the same attribute query with high accuracy by determining the attribute using the search suggestion indicating the total result of the user's search behavior.

  In addition, the second extraction unit 135 is based on the similarity between each content reached from the search result based on each related query extracted by the first extraction unit 134 and the content reached from the search result based on the input query. The same attribute query of the input query is extracted.

  As described above, the providing apparatus 100 according to the embodiment may determine the attribute based on the similarity of the content that is the destination of the search result. Since queries having similar attributes often have similar destinations as search results, the providing apparatus 100 can accurately extract the same attribute query by such a technique.

  Further, the second extraction unit 135 is based on the similarity between the domain of the page reached from the search result by each related query extracted by the first extraction unit 134 and the domain of the page reached from the search result by the input query. Thus, the same attribute query of the input query is extracted.

  As described above, the providing apparatus 100 according to the embodiment may determine the attribute based on the similarity of the domain of the page. As a result, the providing apparatus 100 can compare and determine the similarity of the destinations in more detail, and thus can extract the same attribute query with higher accuracy.

  In addition, the second extraction unit 135 uses the learning model that has learned the relationship with the query having the same or similar attribute as the input query, and among the related queries extracted by the first extraction unit 134, Extract identical attribute queries.

  As described above, the providing apparatus 100 according to the embodiment determines the attribute using the learning model, so that, for example, it is possible to accurately determine the attribute of queries that are difficult to determine based on a search result, a co-occurrence word, or the like. Judgment can be made.

  In addition, when a plurality of input queries are input from the user, the first extraction unit 134 extracts a plurality of related queries related to each input query, and a plurality of related queries with each of the plurality of input queries. The sum of the relevance with the query is calculated, and a predetermined number of related queries are extracted based on the calculated value. The second extraction unit 135 extracts, from the predetermined number of related queries extracted by the first extraction unit 134, the same attribute query having the same or similar attribute as at least one input query of the plurality of input queries.

  As described above, when a plurality of input queries are received, the providing apparatus 100 according to the embodiment extracts a related query based on the sum of the relevance of each input query, or extracts the same attribute query. May be. Accordingly, the providing apparatus 100 can provide, for example, an advertisement or a recommendation having a relationship with any of a plurality of input queries, and thus can provide information more in line with the interest of the user.

  Further, the second extraction unit 135 extracts the same attribute query of the input query first input from the user from the predetermined number of related queries extracted by the first extraction unit 134.

  As described above, the providing apparatus 100 according to the embodiment extracts the same attribute query based on the input query that is input first, which is the query that is most likely to reflect the interest of the user. It is possible to provide information suitable for

  Also, the providing unit 136 provides a user with a recommendation or advertisement related to the same attribute query extracted by the second extraction unit 135.

  As described above, the providing device 100 according to the embodiment provides the user with advertisements and recommendations as information provision. That is, the providing device 100 can provide advertisements and recommendations corresponding to queries having the same or similar attributes as the query input by the user, regardless of the meaning of the query, etc., so that the appeal effect of the information to be provided can be enhanced. it can.

  In addition, the providing apparatus 100 according to the embodiment is based at least on the number of users who input any one of two different queries among arbitrary queries and the number of users who input both the two queries. And a calculation unit 132 that calculates the degree of association between arbitrary queries. The first extraction unit 134 extracts a plurality of related queries related to the input query based on the relevance calculated by the calculation unit 132.

  Thus, since the providing apparatus 100 according to the embodiment calculates the relevance level based on the user's search behavior log, the relevance level is based on the actual interest of the user regardless of the meaning or category of the query itself. Can be calculated. Accordingly, the providing apparatus 100 can calculate the degree of association in which the user's behavior is accurately reflected, such as calculating the degree of association between the queries searched by the user who takes similar behavior.

  As described above, some of the embodiments of the present application have been described in detail with reference to the drawings. However, these are merely examples, and various modifications, including the aspects described in the disclosure section of the invention, based on the knowledge of those skilled in the art, It is possible to implement the present invention in other forms with improvements.

  Further, the providing apparatus 100 described above may be realized by a plurality of server computers, and depending on the function, an external platform or the like may be realized by calling an API (Application Programming Interface) or network computing. Can be changed flexibly.

  Further, “section (module, unit)” described in the claims can be read as “means”, “circuit”, and the like. For example, the acquisition unit can be read as acquisition means or an acquisition circuit.

DESCRIPTION OF SYMBOLS 1 Provision system 10 User terminal 100 Provision apparatus 110 Communication part 120 Storage part 121 Relevance information storage part 122 Attribute information storage part 123 Recommendation storage part 130 Control part 131 Acquisition part 132 Calculation part 133 Reception part 134 1st extraction part 135 2nd Extraction unit 136 Providing unit

Claims (13)

A first extraction unit that extracts a plurality of related queries related to an input query input by a user based on a degree of association between arbitrary queries;
A second extraction unit that extracts the same attribute query that is the query having the same or similar attribute as the input query among the related queries extracted by the first extraction unit;
A providing unit for providing information to the user based on the same attribute query extracted by the second extracting unit;
A providing device comprising: The second extraction unit includes
Extracting the same attribute query based on dictionary data in which an attribute of an arbitrary query is predefined;
The providing device according to claim 1. The second extraction unit includes
Among related queries extracted by the first extraction unit, a related query that is associated with a keyword that is the same as or similar to a keyword that is associated with the input query is extracted as the same attribute query of the input query.
The providing apparatus according to claim 1, wherein: The second extraction unit includes
Among related queries extracted by the first extraction unit, a related query in which a keyword that is the same as or similar to a keyword used as a search suggestion of the input query in a predetermined search engine is used as a search suggestion is the same as the input query. Extract as an attribute query,
The providing device according to claim 3. The second extraction unit includes
The same attribute query of the input query is calculated based on the similarity between each content reached from the search result based on each related query extracted by the first extraction unit and the content reached from the search result based on the input query. Extract,
The providing apparatus according to any one of claims 1 to 4, wherein The second extraction unit includes
The same attribute of the input query based on the similarity between the domain of the page reached from the search result of each related query extracted by the first extraction unit and the domain of the page reached from the search result of the input query Extract queries,
The providing device according to claim 5. The second extraction unit includes
Extracting the same attribute query of the input query from the related queries extracted by the first extraction unit using a learning model that has learned the relationship with a query having the same or similar attribute as the input query,
The providing apparatus according to any one of claims 1 to 6. The first extraction unit includes:
When a plurality of input queries are input from the user, a plurality of related queries related to each input query are extracted, and the degree of relevance between each of the plurality of input queries and the plurality of related queries is added. Calculate, extract a certain number of related queries based on the calculated value,
The second extraction unit includes
Extracting the same attribute query having the same or similar attribute as at least one input query of the plurality of input queries from the predetermined number of related queries extracted by the first extraction unit;
The providing device according to any one of claims 1 to 7. The second extraction unit includes
Extracting the same attribute query of the input query first input from the user from the predetermined number of related queries extracted by the first extraction unit;
The providing apparatus according to claim 8. The providing unit includes:
Providing a recommendation or advertisement related to the same attribute query extracted by the second extraction unit to the user;
The providing device according to any one of claims 1 to 9, wherein Calculate the degree of relevance between arbitrary queries based at least on the number of users who have entered one of two different queries of any query and the number of users who have entered both of the two queries. A calculation unit,
Further comprising
The first extraction unit includes:
Extracting a plurality of related queries related to the input query based on the relevance calculated by the calculation unit;
The providing device according to any one of claims 1 to 10. A providing method executed by a computer,
A first extraction step of extracting a plurality of related queries related to an input query input by a user based on a degree of association between arbitrary queries;
A second extraction step of extracting the same attribute query, which is a query having the same or similar attribute as the input query, from the related queries extracted by the first extraction step;
A providing step of providing information to the user based on the same attribute query extracted by the second extracting step;
The providing method characterized by including. A first extraction procedure for extracting a plurality of related queries related to an input query input by a user based on a degree of association between arbitrary queries;
A second extraction procedure for extracting the same attribute query that is a query having the same or similar attribute as the input query from the related queries extracted by the first extraction procedure;
A providing procedure for providing information to the user based on the same attribute query extracted by the second extracting procedure;
A program for causing a computer to execute.

Images