JP2022126427A - Information processing device, information processing method, and information processing program - Google Patents

Abstract

To provide: an information processing device capable of providing a user with more appropriate information; an information processing method; and an information processing program.SOLUTION: An information processing device 10 comprises: a first acquisition unit which acquires a search query entered by a user on a terminal device T; a learning unit which generates a learning model for converting the search query; and a conversion unit which converts the search query using the learning model. The learning unit learns a relationship between the search query entered by a user and search results selected by the user, from search results corresponding to the search query, and generates a learning model.SELECTED DRAWING: Figure 2

Description

The present invention relates to an information processing device, an information processing method, and an information processing program.

2. Description of the Related Art Information processing technology for providing various information to users is known. A user inputs a search query related to desired information to the information processing device, and the information processing device searches for the information desired by the user based on the input search query.

JP 2019-153271 A

However, depending on the inspection query input by the user, the information requested by the user may not be retrieved.

SUMMARY OF THE INVENTION In view of the above problems, an object of the present invention is to provide appropriate information to users.

In order to solve the above-described problems and achieve the object, the information processing device according to the present disclosure generates a first acquisition unit that acquires a search query input by a user, and a learning model that converts the search query. A learning unit and a conversion unit that converts the search query using the learning model.

According to one aspect of the embodiment, appropriate information can be provided to the user.

FIG. 1 is a diagram illustrating an example of an information processing apparatus according to the first embodiment. FIG. 2 is a diagram illustrating a configuration example of an information processing apparatus according to the first embodiment; FIG. 3 is a diagram showing an example of information registered in the first search query database according to the first embodiment. FIG. 4 is a diagram showing an example of information registered in the second search query database according to the first embodiment. FIG. 5 is a diagram showing an example of information registered in the third search query database according to the first embodiment. FIG. 6 is a diagram showing an example of information registered in the Q&A database according to the first embodiment. FIG. 7 is a diagram illustrating an example of the flow of information processing according to the first embodiment. FIG. 8 is a diagram illustrating an example of a hardware configuration;

Embodiments for implementing an information processing apparatus, an information processing method, and an information processing program according to the present application (hereinafter referred to as "embodiments") will be described in detail below with reference to the drawings. The information processing apparatus, information processing method, and information processing program according to the present application are not limited to this embodiment. Further, in each of the following embodiments, the same parts are denoted by the same reference numerals, and overlapping descriptions are omitted.

[First Embodiment]
[1-1. About an example of an information processing device]
First, an information processing apparatus 10, which is an example of an information processing apparatus according to an embodiment, will be described with reference to FIG.

FIG. 1 is a diagram illustrating an example of an information processing apparatus according to the first embodiment. An information processing device 10 illustrated in FIG. 1 is an information processing device that performs information processing, and is realized by, for example, a server device, a cloud system, or the like. For example, the information processing device 10 communicates with the terminal device used by each user via the network N. FIG. The network N is various wireless communication networks such as 4G (Generation), 5G, LTE (Long Term Evolution), Wifi (registered trademark), or wireless LAN (Local Area Network), or various wired communication networks.

A terminal device QT1 (hereinafter sometimes collectively referred to as a “terminal device T”) is, for example, a PC (Personal Computer), a server device, a smart device such as a smart television, a smart phone, a smart speaker, or a tablet. The terminal device T can communicate with the information processing device 10 via the network N. FIG.

In addition, the terminal device QT1 has a screen such as a liquid crystal display, which has a touch panel function. It may have a function capable of accepting various operations on the content to be processed. In the example shown in FIG. 1, the terminal device QT1 is assumed to be the terminal device used by the questioner Q1.

The information processing apparatus 10 is a server host computer that provides various services to client terminals. The information processing device 10 distributes content to the terminal device QT1. For example, the information processing device 10 includes a portal site, a game information distribution site, a news site, an auction site, a weather forecast site, a shopping site, a finance (stock price) site, a route search site, a map providing site, a travel site, and a restaurant introduction site. , bulletin board sites, web blogs, etc., are distributed to the terminal device T. The information processing apparatus 10 of this embodiment provides an information sharing service that allows users to ask and answer questions.

The information processing apparatus 10 is a server host computer that provides various services to client terminals. For example, the information processing apparatus 10 may be an information transmission service such as an information search service, a SNS (Social Networking Service), a news distribution service, an information compilation service, an Internet encyclopedia service, an Internet dictionary service, a blog service (hereinafter referred to as an information service). .).

Information processing apparatus 10, information processing apparatus 10A, information processing apparatus 10B, and information processing apparatus 10C shown in FIG. The information processing device 10A is a server that provides a service Y1 (for example, a question and answer service), which is one of information services. The information processing device 10B is a server that provides a service Y2 (for example, an Internet encyclopedia service), which is one of information services. The information processing device 10C is a server that provides a service Y3 (for example, an information search service), which is one of information services. In the example shown in FIG. 1, different information services are provided by the four information processing apparatuses, that is, the information processing apparatus 10, the information processing apparatus 10A, the information processing apparatus 10B, and the information processing apparatus 10C. The number of devices is not limited to four, and any other number of information processing devices may be used. That is, one information processing apparatus 10 may be provided, and one information processing apparatus 10 may provide an arbitrary combination of different information services.

For example, in the example shown in FIG. 1, the information processing apparatus 10 causes the learning model to learn the relationship between the search query and the selected search result (step S1). The information processing device 10 receives an input of a search query from the user (step S2). The information processing device 10 converts the input search query (step S3). The information processing device 10 acquires related information based on the converted search query (step S4). The information processing device 10 displays the acquired related information on the user terminal (step S5).

[1-2. Regarding the processing of the information processing device]
Here, depending on the form of the search query input by the user, the information requested by the user may not be retrieved. Therefore, it is required to convert the search query input by the user into a search query that can search for the information that the user desires.

Therefore, the information processing apparatus 10 generates a learning model for converting a search query, and uses the generated learning model to convert the search query input by the user.

For example, processing when a user inputs a search query A, "What is the difference in performance between the tennis rackets of manufacturer A and manufacturer B?" will be described. In this case, the information processing apparatus 10 converts the search query into a keyword search query. That is, the information processing apparatus 10 converts the search query A into a search query of keywords "manufacturer A", "manufacturer B", "tennis racket", and "performance". When the search query A is converted into a keyword search query, the information processing apparatus 10 acquires related information from a plurality of information services based on the keyword search query. After acquiring the related information, the information processing apparatus 10 displays the acquired related information on the terminal device of the user.

As a result, the search query input by the user can be converted into a search query that can acquire information in line with the user's search intent. Therefore, it is possible to provide the user with the information that the user wants to obtain by inputting a search query.

[1-3. About the learning process using the first search query database]
Here, when converting the search query entered by the user, a learning model that learns the relationship between the search query and the search results selected by the user from among the search results searched based on the search query is used. If the search query can be converted by using the above, it is possible to obtain a search query that can obtain search results that match the user's search intent.

Therefore, the information processing apparatus 10 learns the relationship between the search query input by the user and the search result selected by the user from the search results corresponding to the search query, and generates a learning model.

For example, the information processing apparatus 10 associates and registers a search query input by the user and a selected search result, which means a search result selected by the user from among the search results corresponding to the search query. A first search query database 31 is provided. The information processing device 10 uses the data registered in the first search query database 31 to compare the search query input by the user and the search result selected by the user from among the search results corresponding to the search query. Learn relationships and generate learning models. The information processing apparatus 10 acquires a search query input by a user, inputs the acquired search query into a learning model, and obtains an output of a converted search query. After obtaining the output of the converted search query, the information processing apparatus 10 uses the converted search query to obtain related information from a plurality of information services. After acquiring the related information, the information processing apparatus 10 displays the acquired related information on the terminal device of the user.

As a result, the search query input by the user can be converted into a search query that can acquire information in line with the user's search intent. Therefore, it is possible to provide the user with the information that the user wants to obtain by inputting a search query.

[1-4. About the learning process using the second search query database]
Here, when converting the search query entered by the user, a learning model that learns the related information selected by the user from among the related information obtained from multiple information services based on the search query and the search query. If a search query can be converted using this, it is possible to obtain a search query that can acquire related information from a plurality of information services in line with the user's search intent.

Therefore, the information processing apparatus 10 learns and learns the relationship between the search query input by the user and the related information selected by the user from related information acquired from a plurality of information services based on the search query. Generate a model.

For example, the information processing apparatus 10 includes a search query input by the user and selected related information that means related information selected by the user from among related information acquired from a plurality of information services based on the search query. , are registered in association with each other, and a second search query database 32 described later is provided. The information processing apparatus 10 uses the data registered in the second search query database 32 to select a search query input by the user and related information acquired from a plurality of information services based on the search query. It learns the relationship between the selected related information and generates a learning model. The information processing apparatus 10 acquires a search query input by a user, inputs the acquired search query into a learning model, and obtains an output of a converted search query. After obtaining the output of the converted search query, the information processing apparatus 10 uses the converted search query to obtain related information from a plurality of information services. After acquiring the related information, the information processing apparatus 10 displays the acquired related information on the terminal device of the user.

As a result, the search query input by the user can be converted into a search query that can acquire related information in line with the user's search intent. Therefore, it is possible to provide the user with the information that the user wants to obtain by inputting a search query.

[1-5. About the learning process using the third search query database]
Here, when converting the search query entered by the user, the relationship between the search query and the question text selected by the user from among the related question texts obtained from the question answering service based on the search query If a search query can be converted using a learning model that has learned the above, it is possible to obtain a search query that can be obtained from a question answering service that is related to the user's search intention.

Therefore, the information processing device 10 learns the relationship between the search query input by the user and the question selected by the user from the related questions obtained from the question answering service based on the search query, Generate a learning model.

For example, the information processing apparatus 10 selects a question sentence selected by the user from among the search query input by the user and the related question sentences obtained from the question answering service based on the search query. and a third search query database 33, which will be described later, in which are registered in association with each other. The information processing apparatus 10 also includes a Q&A database 34, which will be described later, in which a question text input by a user and a set of answer texts input by other users to the question text are associated and registered. . Using the data registered in the third search query database 33, the information processing apparatus 10 selects the user from among the search query input by the user and related question sentences obtained from the question answering service based on the search query. learns the relationship between the question sentences selected by and generates a learning model. The information processing apparatus 10 acquires a search query input by a user, inputs the acquired search query into a learning model, and obtains an output of a converted search query. After obtaining the output of the converted search query, the information processing apparatus 10 uses the converted search query to obtain related question sentences from the question answering service. The information processing apparatus 10 displays the related question on the terminal device of the user when the related question is acquired. In addition, the information processing apparatus 10 adds a link for accessing a set of answer sentences input by other users to the relevant question sentence, and displays the linked question sentence on the terminal device of the user.

As a result, the search query input by the user can be converted into a search query that can acquire a question sentence in line with the user's search intention. Therefore, it is possible to provide the user with the information that the user wants to obtain by inputting a search query.

[1-6. About conversion process]
Here, the information processing apparatus 10 converts a natural sentence search query into a keyword search query using a learning model for converting a search query.

For example, processing when a user inputs a search query A "What is the difference in performance between the tennis rackets of manufacturer A and manufacturer B?" will be described. In this case, the information processing apparatus 10 converts the search query A into a search query of keywords "manufacturer A", "manufacturer B", "tennis racket", and "performance". The information processing apparatus 10 acquires related information from a plurality of information services based on the converted keyword search query. After acquiring the related information, the information processing apparatus 10 displays the acquired related information on the terminal device of the user.

As a result, a natural sentence search query input by the user can be converted into a keyword search query that can acquire information in line with the user's search intent. Therefore, it is possible to provide the user with the information that the user wants to obtain by inputting a search query in natural sentences.

[1-7. Display of related information]
Here, when the user inputs a search query, in addition to the search results of the normal search, related information can be obtained from a plurality of information services based on the search query and displayed on the user's terminal device. It is thought that it is possible to activate the use of multiple information services.

Therefore, the information processing apparatus 10 acquires related information from a plurality of information services based on the converted search query.

For example, processing when a user inputs a search query A "What is the difference in performance between the tennis rackets of manufacturer A and manufacturer B?" will be described. In this case, the information processing apparatus 10 converts the search query A into a search query of keywords "manufacturer A", "manufacturer B", "tennis racket", and "performance". The information processing apparatus 10 acquires related information from a plurality of information services based on the converted keyword search query. The information processing apparatus 10 acquires, for example, information on "tennis rackets" of "manufacturer A" and "manufacturer B" from a shopping site as related information. Further, the information processing apparatus 10 acquires, for example, information on "tennis rackets" of "manufacturer A" and "manufacturer B" from an auction site as related information. The information processing apparatus 10 also acquires, for example, information related to "manufacturer A", "manufacturer B", "tennis racket", and "performance" from the bulletin board site as related information. After acquiring the related information, the information processing apparatus 10 displays the acquired related information together with other search results on the terminal device of the user.

As a result, it is possible to convert the search query input by the user, acquire related information from a plurality of information services based on the converted search query, and display the acquired related information on the user's terminal device. . Therefore, the information requested by the user can be appropriately provided. Also, it is possible to activate the use of a plurality of information services.

[1-8. Regarding the display of the question text]
Here, when the user inputs a search query, in addition to the search results of the normal search, a question sentence related to the search query can be acquired from the question and answer service and displayed on the user's terminal device. If possible, it is thought that the information requested by the user can be provided. As a result, it is possible to activate the use of the question answering service registered in the portal site.

Therefore, the information processing apparatus 10 acquires related question sentences from the question answering service based on the converted search query.

For example, processing when a user inputs a search query A "What is the difference in performance between the tennis rackets of manufacturer A and manufacturer B?" will be described. In this case, the information processing apparatus 10 converts the search query A into keyword search queries B “manufacturer A”, “manufacturer B”, “tennis racket”, and “performance”. The information processing apparatus 10 acquires a related question sentence from the question answering service based on the converted keyword search query B. FIG. For example, a question sentence A containing all of the keyword search query B is posted by a user to a question answering service, "What is the difference in performance between the tennis rackets of manufacturer A and manufacturer B?" It is assumed that an answer has been posted by a user of In this case, the information processing device 10 acquires the question sentence A from the question answering service. After obtaining the question text A, the information processing apparatus 10 displays the obtained question text A on the terminal device of the user together with other search results.

As a result, it is possible to convert the search query entered by the user, acquire related questions from the question and answer service based on the converted search queries, and display the acquired questions on the user's terminal device. can. Therefore, the information requested by the user can be appropriately provided.

[1-9. Acquisition of similar question sentences using clusters of distributed representations of question sentences]
Here, when searching question texts, not only question texts that include all of the keywords of the search query, but also question texts that do not include all of the keywords of the search query but have similar meanings are displayed. By referring to the related question text, the user can obtain information that matches the search intention of the input search query.

Therefore, the information processing apparatus 10 clusters the distributed representations corresponding to the character information of the question text based on the degree of similarity, and clusters the distributed representations corresponding to the character information of the search query after conversion and the distributed representation corresponding to the character information of the question sentence. , and based on the calculated similarity, a cluster of question sentences related to the search query is selected, and the question sentences included in the selected cluster are obtained.

For example, in a question-answering service, a user asks question sentence B, "What is the difference in performance between the tennis rackets of manufacturer C and manufacturer D?" It is assumed that the user has input a question sentence D, "Do you have any tennis racket manufacturers?" Assume that another user has input an answer to question B, question C, and question D. In this case, the information processing apparatus 10 calculates distributed representations for the question text B, the question text C, and the question text D. FIG. Here, since the question sentence B, the question sentence C, and the question sentence D are similar in meaning, the values of the distributed expressions are calculated to be close to each other. The information processing apparatus 10 classifies the question text B, the question text C, and the question text D into the same cluster A after the distributed representations of the question text B, the question text C, and the question text D are calculated. Here, it is assumed that the information processing apparatus 10 acquires the same search query A as the above-described question text A from the user. After acquiring the search query A, the information processing apparatus 10 calculates a distributed representation of the acquired search query A. FIG. Since the question text A is similar in meaning to the question text B, the question text C, and the question text D, the distributed representation of the same search query A as the question text A is the question text B, the question text C, A value close to the distributed representation of the question sentence D is calculated. When the distributed representation of the search query A is calculated, the information processing apparatus 10 acquires the question text B from the question texts classified into the cluster A, for example. The number of question sentences acquired from cluster A is not limited to one, and any other number of question sentences may be acquired. After obtaining the question B from the cluster A, the information processing apparatus 10 displays the question B together with other search results on the terminal device of the user.

This makes it possible to provide the user with information in line with the search intent of the search query when the user inputs the search query. Therefore, the information requested by the user can be appropriately provided.

[2. Configuration of Information Processing Device]
An example of the functional configuration of the information processing apparatus 10 described above will be described below. Note that the following description shows an example of the functional configuration of the information processing apparatus 10 that executes information processing at the information service site. FIG. 2 is a diagram illustrating a configuration example of an information processing apparatus according to the first embodiment; As shown in FIG. 2 , the information processing device 10 has a communication section 20 , a storage section 30 and a control section 40 .

The communication unit 20 is realized by, for example, a NIC (Network Interface Card) or the like. The communication unit 20 is connected to the network N by wire or wirelessly, and transmits and receives information to and from the terminal device T. FIG.

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.

The storage unit 30 also stores a first search query database 31, a second search query database 32, a third search query database 33, a Q&A database 34, a learning model 35, and a distributed representation database 36. .

In the first search query database 31, the search query input by the user and the information of the search result selected by the user with respect to the search result searched based on the search query are registered. For example, FIG. 3 is a diagram showing an example of information registered in the first search query database 31 according to the embodiment. As shown in FIG. 3, in the first search query database 31, information having items such as "user ID", "search query", and "selected search result" is registered.

Here, "user ID" is an identifier for identifying a user. A "search query" is information indicating a search query input by a user. The "selected search result" is information indicating the search result selected by the user from the search results obtained based on the search query input by the user. In addition, the search query is uniformly obtained from all users without specifying the user, and the selected search result indicating the search result selected by the user for the search result searched based on the obtained search query is obtained. Then, the acquired search queries and unselected search queries may be registered in the first search query database 31 .

For example, in the example shown in FIG. 3, the first search query database 31 stores a user ID "U1", a search query "search query #U1-1", and a selected search result "selected search result #U1-1". , is registered. Such information is obtained by inputting a search query indicated by the search query "search query #U1-1" by the user indicated by the user ID "U1", and obtaining the selected search result "selected search result #U1-1". indicates that the information indicated by is selected.

In the example shown in FIG. 3, the information registered in the first search query database 31 is described using conceptual expressions such as search query #U1 and selected search result #U1-1. , character information as a search query input by the user, and character information of selected search results indicating the search results selected by the user for the search results obtained based on the search query are registered. In addition to the information shown in FIG. 3 , any other information related to search queries may be registered in the first search query database 31 .

In the second search query database 32, a search query input by the user, related information selected by the user from related information acquired from a plurality of information services based on the search query, and related information are acquired. The name of the information service is registered. For example, FIG. 4 is a diagram showing an example of information registered in the second search query database 32 according to the first embodiment. As shown in FIG. 4, in the second search query database 32, information having items such as "user ID", "search query", "selected related information", and "information service name" is registered.

For example, in the example shown in FIG. 4, information such as user ID "U1", search query "search query #U1-1", and information service name "information service A" are associated and registered. Such information is obtained by inputting a search query indicated by a search query "search query #U1-1" by a user indicated by a user ID "U1", and obtaining non-selected related information "non-selected related information #U1-1". indicates that the related information indicated by is selected and that the selected related information is acquired from the information service indicated by the information service name "information service A".

In the example shown in FIG. 4, conceptual expressions such as "search query #U1-1" and "information service A" are described. Character information of related information selected by the user from related information acquired from multiple information services based on a search query, and character information of the name of the information service from which the related information selected by the user is acquired are registered. It will happen. In addition to the information shown in FIG. 4 , any other information related to search queries may be registered in the second search query database 32 .

In the third search query database 33, search queries input by the user and questions selected by the user from among the questions obtained from the question answering service based on the search query are registered. For example, FIG. 5 is a diagram showing an example of information registered in the third search query database 33 according to the first embodiment. As shown in FIG. 5, in the third search query database 33, information having items such as "user ID", "search query", and "selected question" is registered.

For example, in the example shown in FIG. 5, information such as a user ID “U1”, a search query “search query #U1-1”, and a selected question “selected question #U1-1” are registered in association with each other. there is Such information is obtained by inputting a search query indicated by a search query "search query #U1-1" by a user indicated by a user ID "U1", and obtaining a non-selected question "non-selected question #U1-1". indicates that the question sentence indicated by is selected.

In the example shown in FIG. 5, conceptual expressions such as "search query #U1-1" and "selected question sentence #U1-1" are described, but in reality, the search query input by the user and the character information of the question sentence selected by the user from the related question sentences obtained from the question answering service based on the search query are registered. In addition to the information shown in FIG. 5, any other information related to search queries may be registered in the third search query database 33 .

In the Q&A database 34, a set of questions input by the user, a set of answers input by other users to the questions, and questions from the set of answers input by other users are stored. Information related to the answer selected as the best answer by the input user is registered. For example, FIG. 6 is a diagram showing an example of information registered in the Q&A database according to the first embodiment. As shown in FIG. 6, in the Q&A database 34, information having items such as "question ID", "question sentence", "answer sentence set", and "selection information" is registered. A set of answer sentences is information in which answer sentences input by one or more other users to a question sentence input by a user are grouped together.

In the example shown in FIG. 6, information such as question ID "Q1", question sentence "question sentence #1", answer sentence set "answer sentence set #1", and selection information "answer sentence #1-1" are registered in association with each other. It is As for such information, a question ID "Q1" is given to the question text input by the user, the question text input by the user is registered in the question text "question #1", and the question text input by the user is registered. Answer sentences shown in the answer sentence set "answer sentence set #1" are input by other users to the question sentences, and the answer sentences indicated by the selection information "answer sentence #1-1" are input by the user who has entered the question sentences. selected as the best answer.

In addition, in the example shown in FIG. 6, conceptual expressions such as "question sentence #1", "answer sentence #1", and "answer sentence #1-1" are described. Character information of the question text entered by the user, character information of a set of answers entered by other users to the question text entered by the user, and character information of the answer text selected by the user who entered the question text to be registered. In addition to the information shown in FIG. 6, for example, the Q&A database 34 stores the user ID of the user who entered the question and the user ID of the user who entered the answer to the entered question. Arbitrary information regarding various question sentences and answer sentences such as IDs may be registered.

The learning model 35 stores an unlearned learning model and a learned learning model. As the learned learning model, a learning model generated by execution of learning processing by the learning unit 42, which will be described later, is stored. When the learning unit 42 (to be described later) executes the learning process, the learning process is executed using the unlearned learning model stored in the learning model 35 .

In the distributed representation database 36, the question text input by the user, the distributed representation of the question text, and the cluster classified into the question text are registered in association with each other.

The control unit 40 is, for example, a controller, and various programs stored in the storage device of the information processing apparatus 10 are executed using the RAM as a work area by means of a CPU (Central Processing Unit), an MPU (Micro Processing Unit), or the like. It is realized by being executed. Also, the control unit 40 is a controller, and may be implemented by an integrated circuit such as an ASIC (Application Specific Circuit) or an FPGA (Field Programmable Gate Array).

As shown in FIG. 2 , the control unit 40 has an acquisition unit 41 , a learning unit 42 , a conversion unit 43 , a calculation unit 44 , a classification unit 45 , a reception unit 46 and a provision unit 47 .

Acquisition unit 41 includes first acquisition unit 411 , second acquisition unit 412 , and third acquisition unit 413 . The first acquisition unit 411 acquires a search query input by the user. For example, if the user inputs a search query A "What is the performance difference between the tennis rackets of manufacturer A and manufacturer B?" A search query input to the terminal device T via the communication unit 20 is acquired.

The second acquisition unit 412 acquires related information from a plurality of information services based on a search query that is converted by the conversion unit 43, which will be described later, from the search query input by the user. For example, it is assumed that the user inputs a search query A to the terminal device T, "What is the difference in performance between the tennis rackets of manufacturer A and manufacturer B?" In this case, the first acquisition unit 411 acquires the search query A input by the user, and converts the search query A acquired by the conversion unit 43 (to be described later) into search queries B "manufacturer A", "manufacturer B", and "manufacturer A". "Tennis Racket", translates to "Performance". The second acquisition unit 412 acquires related information from a plurality of information services based on the search query B converted by the conversion unit 43 . The second acquisition unit 412 acquires, for example, the sales information regarding the “tennis racket” of the “manufacturer A” and the “manufacturer B” from the shopping site as related information.

The third acquisition unit 413 selects a cluster of question sentences based on the calculation result of the distributed representation of the search query input by the user, and acquires the question sentences included in the selected cluster. That is, when the calculation unit 44, which will be described later, calculates the distributed representation of the search query input by the user, the third acquisition unit 413 obtains the distributed representation of the search query input by the user and the distributed representation of the cluster of question sentences. A similarity is calculated, and a cluster of question sentences with a high similarity is selected. When a cluster of question sentences with a high degree of similarity is selected, the third acquisition unit 413 acquires question sentences included in the selected cluster.

The learning unit 42 generates a learning model for converting search queries. For example, the learning unit 42 learns the relationship between the search queries registered in the first search query database 31 and the selected search results, and generates a learning model for converting the search queries. In this case, for example, a learning model may be generated that outputs a search query input by the user when a natural sentence included in the character information of the selected search result is input. Also, for example, the learning unit 42 learns the relationship between the search queries registered in the second search query database 32 and the selected related information, and generates a learning model for converting the search queries. In this case, for example, a learning model may be generated that outputs a search query input by the user when a natural sentence included in the character information of the selected related information is input. Also, for example, the learning unit 42 learns the relationship between the search queries registered in the third search query database 33 and the selected question sentences, and generates a learning model for converting the search queries. In this case, for example, a learning model may be generated that outputs a search query input by the user when a natural sentence included in the character information of the selected question is input. That is, the learning model transforms the search query according to the user's behavior after entering the search query. In other words, the learning model is learned so as to convert the search query according to the search intention estimated from the user's behavior after inputting the search query.

The learning unit 42 generates a trained learning model by end-to-end deep learning. End-to-end deep learning replaces machine learning systems that required multiple stages of processing from input data to output results as a single large neural network with multiple layers and modules that perform various processes. Learn by replacing with That is, end-to-end deep learning directly learns the correspondence between inputs and outputs. By using end-to-end deep learning, there is no need to prepare an intermediate structure specialized for a specific function, and it is necessary to consider the problem that problems in the upstream process that occur in natural language processing are postponed to the downstream process. is gone.

The conversion unit 43 converts the search query using the learning model. For example, the conversion unit 43 converts the search input by the user into the learning model generated by the learning unit 42 learning the relationship between the search queries registered in the first search query database 31 and the selected search results. By inputting a query, we get the output of the transformed search query. Further, for example, the conversion unit 43 allows the learning unit 42 to learn the relationship between the search query registered in the second search query database 32 and the selected related information, and the user inputs By inputting the search query, we get the output of the converted search query. Further, for example, the conversion unit 43 allows the learning unit 42 to learn the relationship between the search queries registered in the third search query database 33 and the selected question sentences, and the user inputs to the learning model generated. By inputting the converted search query, the output of the converted search query is obtained.

The calculator 44 calculates a distributed representation corresponding to character information. For example, the calculation unit 44 performs morphological analysis on the question sentence, segments it into morphemes, and converts the segmented morphemes into distributed representations. Word2vec, for example, can be used to convert words into distributed representation. After converting all the words forming the question sentence into the distributed representation, the distributed representation of the question sentence can be obtained by, for example, calculating the average value of the distributed representations of all the words forming the question sentence. After calculating the distributed representation of the question sentence, the calculation unit 44 registers the distributed representation of the question sentence in the distributed representation database 36 in association with the question sentence. Even if the search query is a natural sentence, a distributed representation of the search query in natural sentence can be obtained by executing similar processing.

The classification unit 45 clusters the distributed representations corresponding to the character information of the question sentence based on the degree of similarity. For example, the classification unit 45 calculates the cosine similarity between the distributed representation corresponding to the character information of the question and the distributed representation corresponding to the character information of the other question. Note that the classification unit 45 may calculate the similarity between distributed representations based on any index, not limited to the cosine similarity, as long as the index is applicable as a distance measure between vectors. For example, the classification unit 45 may calculate the value of a predetermined distance function such as the Euclidean distance between distributed representations, the distance in a non-Euclidean space such as a hyperbolic space, the Manhattan distance, the Mahalanobis distance, or the like.

After calculating the similarity of the distributed expressions corresponding to the character information of the question, the classification unit 45 clusters the distributed expressions of the question based on the similarity of the distributed expressions. After clustering the distributed representations of the question sentences, the classification unit 45 registers the clusters of the distributed representations of the question sentences in the distributed representation database 36 in association with the question sentences and the distributed representations of the question sentences. K-means clustering, for example, can be used for clustering distributed representations of question sentences. K-means clustering classifies data into a given number of clusters (K) using cluster mean values.

The reception unit 46 receives registration of questions and answers from users. For example, upon receiving a question sentence from the user's terminal device, the accepting unit 46 registers the accepted sentence in the Q&A database 34 as a question sentence. Further, when accepting sentences as answers from terminal devices used by respondents, the accepting unit 46 associates the accepted sentences with the question ID of the corresponding question sentence and registers them in the Q&A database 34 . As a result, the Q&A database 34 can be updated by receiving new questions and answers posted by users.

The providing unit 47 provides at least either the related information acquired by the second acquiring unit 412 or the question text acquired by the third acquiring unit 413 to the user who has entered the search query. That is, the providing unit 47 may display the related information acquired by the second acquiring unit 412 or may display the question text acquired by the third acquiring unit 413 . Further, the providing unit 47 may display both the related information acquired by the second acquiring unit 412 and the question text acquired by the third acquiring unit 413 on the terminal device of the user.

[3. Processing procedure]
Next, the procedure of information processing by the information processing apparatus 10 according to the first embodiment will be described with reference to FIG. FIG. 7 is a flowchart showing an example of the flow of information processing according to the first embodiment. For example, the information processing device 10 acquires a search query input by the user (step S101). The information processing apparatus 10 then converts the acquired search query (step S102). Then, the information processing apparatus 10 acquires related information from a plurality of information services (step S103). Then, the information processing apparatus 10 displays the acquired related information on the terminal device of the user who entered the search query (step S104).

[4. Hardware configuration]
Further, the information processing apparatus 10 according to the above-described embodiment is implemented by a computer 1000 configured as shown in FIG. 8, for example. FIG. 8 is a diagram illustrating an example of a hardware configuration; A computer 1000 is connected to an output device 1010 and an input device 1020, and an arithmetic device 1030, a primary storage device 1040, a secondary storage device 1050, an output IF (Interface) 1060, an input IF 1070, and a network IF 1080 are connected by a bus 1090. have

The arithmetic device 1030 operates based on programs stored in the primary storage device 1040 and the secondary storage device 1050, programs read from the input device 1020, and the like, and executes various processes. The primary storage device 1040 is a memory device such as a RAM that temporarily stores data used by the arithmetic device 1030 for various calculations. The secondary storage device 1050 is a storage device in which data used for various calculations by the arithmetic device 1030 and various databases are registered, and includes a ROM (Read Only Memory), HDD (Hard Disk Drive), flash memory, and the like. It is realized by

The output IF 1060 is an interface for transmitting information to be output to the output device 1010 that outputs various types of information such as a monitor and a printer. It is realized by a connector conforming to a standard such as HDMI (registered trademark) (High Definition Multimedia Interface). Also, the input IF 1070 is an interface for receiving information from various input devices 1020 such as a mouse, keyboard, scanner, etc., and is realized by, for example, USB.

Note that the input device 1020 includes, for example, optical recording media such as CDs (Compact Discs), DVDs (Digital Versatile Discs), PDs (Phase change rewritable discs), magneto-optical recording media such as MOs (Magneto-Optical discs), and tapes. It may be a device that reads information from a medium, a magnetic recording medium, a semiconductor memory, or the like. Also, the input device 1020 may be an external storage medium such as a USB memory.

Network IF 1080 receives data from other devices via network N and sends the data to arithmetic device 1030, and also transmits data generated by arithmetic device 1030 via network N to other devices.

The arithmetic device 1030 controls the output device 1010 and the input device 1020 via the output IF 1060 and the input IF 1070 . For example, arithmetic device 1030 loads a program from input device 1020 or secondary storage device 1050 onto primary storage device 1040 and executes the loaded program.

For example, when the computer 1000 functions as the information processing device 10 , the arithmetic device 1030 of the computer 1000 implements the functions of the control unit 40 by executing programs loaded on the primary storage device 1040 .

[5. effect〕
The information processing apparatus 10 according to the present disclosure includes a first acquisition unit 411 that acquires a search query input by a user, a learning unit 42 that generates a learning model for converting the search query, and a search query using the learning model. and a converting unit 43 for converting.

According to this configuration, since the search query input by the user is converted using the learning model, appropriate information can be provided to the user.

The learning unit 42 also learns the relationship between the search query input by the user and the search result selected by the user from the search results corresponding to the search query, and generates a learning model.

According to this configuration, it is possible to generate a learning model that can convert a search query input by a user into a search query that can obtain search results that match the user's search intent. Therefore, appropriate information can be provided to the user.

In addition, the learning unit 42 learns the relationship between the search query input by the user and the related information selected by the user from related information acquired from a plurality of information services based on the search query, and learns the learning model. to generate

According to this configuration, it is possible to generate a learning model that can convert a search query input by the user into a search query that can acquire related information in line with the user's search intent. Therefore, appropriate information can be provided to the user.

In addition, the learning unit 42 learns the relationship between the search query input by the user and the question text selected by the user from the related question texts obtained from the question answering service based on the search query, and learns the relationship. Generate a model.

According to this configuration, it is possible to generate a learning model that can convert a search query input by a user into a search query that can acquire a question sentence that matches the user's search intention. Therefore, appropriate information can be provided to the user.

The conversion unit 43 also converts a natural sentence search query into a keyword search query using a learning model.

According to this configuration, a natural sentence search query input by the user can be converted into a keyword search query. Therefore, appropriate information can be provided to the user.

The information processing apparatus 10 according to the present disclosure also includes a second acquisition unit 412 that acquires related information from among a plurality of information services based on the search query converted by the conversion unit 43 .

According to this configuration, it is possible to convert a search query input by the user into a search query that enables searching for related information in line with the user's search intent, and to acquire related information from a plurality of information services. Therefore, appropriate information can be provided to the user.

Also, the second acquisition unit 412 acquires a question sentence as related information from the question answering service as the information service based on the search query converted by the conversion unit 43 .

According to this configuration, the search query input by the user is converted into a search query that can search for a question sentence in line with the user's search intention, and the question sentence is acquired as related information from the question answer service as an information service. can. Therefore, appropriate information can be provided to the user.

Further, the information processing apparatus 10 according to the present disclosure includes a calculation unit 44 that calculates distributed representations corresponding to text information, a classification unit 45 that clusters question sentences based on the similarity of the distributed representations corresponding to question sentences, A third acquisition unit 413 that selects a cluster of question sentences based on the calculation result of the distributed expression of the search query input by the user and acquires the question sentences included in the selected cluster.

According to this configuration, it is possible to select a cluster of question sentences based on the calculation result of the distributed representation of the search query input by the user, and acquire the question sentences included in the selected cluster. Therefore, appropriate information can be provided to the user.

An information processing method according to the present disclosure includes steps of acquiring a search query input by a user, converting the search query, and acquiring related information from a plurality of information services based on the converted search query. , and displaying the obtained relevant information.

According to this configuration, it is possible to convert the search query input by the user into a search query capable of searching for information in line with the user's search intention, and acquire related information from among a plurality of information services. Therefore, appropriate information can be provided to the user.

An information processing program according to the present disclosure includes a step of acquiring a search query entered by a user, a step of converting the search query, and a step of acquiring related information from a plurality of information services based on the converted search query. and causing the computer to display the acquired relevant information.

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

Also, the above-mentioned "section, module, unit" can be read as "means" or "circuit". For example, the acquisition unit can be read as acquisition means or an acquisition circuit.

10 information processing device 20 communication unit 30 storage unit 31 first search query database 32 second search query database 33 third search query database 34 Q&A database 35 learning model 36 distributed representation database 40 control unit 41 acquisition unit 411 first acquisition unit 412 Second acquisition unit 413 Third acquisition unit 42 Learning unit 43 Conversion unit 44 Calculation unit 45 Classification unit 46 Reception unit 47 Provider N Network T Terminal device

Claims (10)

a first acquisition unit that acquires a search query input by a user;
a learning unit that generates a learning model that converts the search query according to the behavior of the user;
a conversion unit that converts the search query using the learning model;
Information processing device. The learning unit learns the relationship between the search query input by the user and the search result selected by the user from among the search results corresponding to the search query, and generates a learning model.
The information processing device according to claim 1 . The learning unit learns the relationship between a search query input by a user and related information selected by the user from related information acquired from a plurality of information services based on the search query, and generates a learning model. do,
The information processing device according to claim 1 . The learning unit learns the relationship between the search query input by the user and the question text selected by the user from the related question texts obtained from the question answering service based on the search query, and creates a learning model. generate,
The information processing device according to claim 1 . The conversion unit converts a natural sentence search query into a keyword search query using the learning model.
The information processing apparatus according to any one of claims 1 to 4. a second acquisition unit that acquires related information from among a plurality of information services based on the search query converted by the conversion unit;
The information processing apparatus according to any one of claims 1 to 4. The second acquisition unit acquires a question sentence as related information from a question and answer service as an information service based on the search query converted by the conversion unit.
The information processing device according to claim 6 . a computing unit that computes a distributed representation corresponding to character information;
a classifier that clusters the question sentences based on the similarity of the distributed representations corresponding to the question sentences;
a third acquisition unit that selects a cluster of question sentences based on the calculation result of the distributed expression of the search query input by the user and acquires the question sentences included in the selected cluster;
The information processing apparatus according to any one of claims 1 to 7. obtaining a search query entered by a user;
transforming the search query;
obtaining relevant information from a plurality of information services based on the transformed search query;
displaying the relevant information obtained;
Information processing method including. obtaining a search query entered by a user;
transforming the search query;
obtaining relevant information from a plurality of information services based on the transformed search query;
displaying the relevant information obtained;
An information processing program that causes a computer to execute

Images