Information processing system, program, and information processing method

Abstract

To reduce time and effort required for a user to acquire desired information in an information processing system that conducts a search.SOLUTION: An information processing system extracts, from a search query input by a user, keywords associated with words included in the search query, and presents the keywords to the user in order based on the degree of association. The degree of association is calculated on the basis of the ratio of information including the words associated with the keywords to information registered in the information processing system. The user selects, from the keywords, one or more keywords representing the details of the search query. The information processing system searches for information associated with the keywords selected by the user and presents a result of the search to the user.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to an information processing system, a program, and an information processing method for a user to search for information. [Background technology]

[0002] Recently, various services have been using systems to search for FAQs (frequently asked questions). By using these systems, users can solve problems without using a call center before contacting the call center. For business owners who set up call centers, these systems are widely adopted because they can reduce the workload of call center operations and the number of operators.

[0003] For example, Patent Document 1 below discloses a system for searching FAQs from natural language. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Laid-Open No. 2015-36945 Summary of the Invention [Problem to be solved by the invention]

[0005] However, the invention of Patent Document 1 had the problem that the user could not tell what combination of keywords was used to search and what results were displayed. Also, if the input query contained unnecessary information and the user wanted to perform a search that did not include part of the input content, the user had to start over from the beginning by inputting the search content.

[0006] The present invention has been made in consideration of the above problems, and an object of the present invention is to provide a new and improved information processing system, program, and information processing method that can reduce the amount of effort required for a user to use a search system to obtain desired information. [Means for solving the problem]

[0007] In order to solve the above problem, according to one aspect of the present invention, An information processing system for searching registered information, An input receiving unit that receives a search query, a keyword extracting unit that extracts keywords related to words included in the search query, and a search engine that uses the keywords extracted by the keyword extracting unit. , love an evaluation value calculation unit that calculates an evaluation value for each keyword based on the proportion of information that includes a word associated with the keyword among the information; a display control unit that performs display control to display keywords related to the terminal in an order based on the evaluation value; and a search unit that searches for information related to a selected keyword among the keywords that are display-controlled by the display control unit. ,table The display control unit 、 Display control is performed to display information related to the selected keyword on the device. The evaluation value calculation unit calculates the ratio of information that includes a word associated with the selected keyword for each category to which the information belongs, and determines the sum of the ratios for each category as the evaluation value. An information processing system is provided.

[0009] Also, According to one aspect of the present invention, there is provided an information processing system for searching registered information, the system comprising: an input receiving unit for receiving a search query; a keyword extraction unit for extracting keywords related to words included in the search query; an evaluation value calculation unit for calculating an evaluation value for each keyword extracted by the keyword extraction unit based on a proportion of information containing words associated with the keyword; a display control unit for performing display control to display the related keywords on a terminal in an order based on the evaluation values; and a search unit for searching for information related to a selected keyword from among the keywords display-controlled by the display control unit, wherein the display control unit performs display control to display information related to the selected keyword on the terminal; The evaluation value calculation unit calculates the evaluation values ​​of the keywords registered in the information processing system when a keyword is registered, deleted, or changed in the information processing system, or when information is registered, deleted, or changed, and calculates the evaluation values ​​of each keyword and An information processing system is provided.

[0010] The display control unit may also perform display control so as to highlight keywords whose evaluation values ​​are equal to or greater than a threshold value or keywords that have been set as priorities.

[0011] Also, According to one aspect of the present invention, there is provided an information processing system for searching registered information, comprising: an input receiving unit for receiving a search query; a keyword extraction unit for extracting keywords related to words included in the search query; an evaluation value calculation unit for calculating an evaluation value for each keyword extracted by the keyword extraction unit based on a proportion of information containing words associated with the keyword; a display control unit for performing display control to display keywords related to a terminal in an order based on the evaluation values; and a search unit for searching for information related to a selected keyword from the keywords display-controlled by the display control unit, wherein the display control unit performs display control to display information related to the selected keyword on the terminal, and performs display control to highlight and display keywords having an evaluation value equal to or greater than a threshold value or keywords that have been set as a priority; The search unit searches for information including keywords with evaluation values ​​above a threshold or keywords that have been prioritized in addition to the selected keywords. An information processing system is provided.

[0012] The display control unit may also include a relevance calculation unit that calculates the relevance between the selected keyword and the information searched by the search unit, and the display control unit may determine the order in which to display the information based on the relevance.

[0013] Also, According to one aspect of the present invention, there is provided an information processing system for searching registered information, comprising: an input receiving unit for receiving a search query; a keyword extraction unit for extracting keywords related to words included in the search query; an evaluation value calculation unit for calculating an evaluation value for each keyword extracted by the keyword extraction unit based on the proportion of information containing words associated with the keyword; a display control unit for performing display control to display related keywords on a terminal in order based on the evaluation value; a search unit for searching for information related to a selected keyword from among the keywords displayed and controlled by the display control unit; and an association calculation unit for calculating the association between the selected keyword and the information searched for by the search unit. a feedback receiving unit that receives a response as to whether the information whose display has been controlled by the display control unit satisfies the requirements of the search query; the display control unit determines an order in which to display information based on the relevance, and performs display control to display information related to the selected keyword on the terminal; The relevance calculation unit calculates the relevance based on past responses in the feedback reception unit. An information processing system is provided.

[0014] The relevance calculation unit may also calculate the relevance using a model that has learned the relationship between the keyword and the relevance of past answers using machine learning.

[0015] The input receiving unit may also receive the search query by voice input.

[0016] The registered information may also be in natural language in which questions and answers are associated with each other.

[0017] In order to solve the above problems, according to another aspect of the present invention, Search for registered information The computer is configured to function as an input receiving unit that receives a search query, a keyword extraction unit that extracts keywords related to words included in the search query, an evaluation value calculation unit that calculates an evaluation value for each keyword extracted by the keyword extraction unit based on the proportion of information that contains words associated with the keyword, a display control unit that controls display to display related keywords on a terminal in order based on the evaluation value, and a search unit that searches for information related to a selected keyword from among the keywords that are display-controlled by the display control unit. At the same time, Display control unit is selectedThe display control is performed so that information related to the selected keyword is displayed on the terminal, and the evaluation value calculation unit calculates the ratio of information containing words related to the selected keyword for each category to which the information belongs, and the sum of the ratios for each category is used as the evaluation value. Make it work , the program is provided.

[0018] In order to solve the above problems, according to another aspect of the present invention, An information processing method for an information processing system that searches for registered information, comprising: A method for calculating a search result by an evaluation value, comprising: receiving a search query input to a terminal by an input receiving unit; extracting keywords related to a word included in the received search query by a keyword extracting unit; By the keyword extraction section The extracted keywords , love calculating an evaluation value for each keyword based on a ratio of information that includes a word associated with the keyword among the information; 、 The method includes a step of performing display control to display keywords related to the terminal in an order based on the evaluation value, a step of a search unit searching for information related to a selected keyword among the keywords displayed on the terminal by the display control, and a step of a display control unit performing display control to display information related to the selected keyword on the terminal. The evaluation value calculation unit calculates the percentage of information that contains words associated with the selected keyword for each category to which the information belongs, and sets the sum of the percentages for each category as the evaluation value. Processing is performed by a computer in an information processing system. is executed A method for processing information is provided. [Effects of the Invention]

[0019] According to the present invention described above, it is possible for a user to reduce the amount of work required to obtain desired information using a search system. [Brief explanation of the drawings]

[0020] [Figure 1] 1 is an explanatory diagram illustrating an overview of an information processing system according to an embodiment of the present invention. [Figure 2] FIG. 2 is an explanatory diagram illustrating an example of the configuration of a server according to the embodiment. [Figure 3]2 is an explanatory diagram showing functional blocks of a Web server 110 according to the embodiment. FIG. [Figure 4] FIG. 2 is an explanatory diagram showing functional blocks of a keyword evaluation server 120 according to the embodiment. [Figure 5] FIG. 10 is an explanatory diagram showing an example of registration of keyword information according to the embodiment; [Figure 6] FIG. 2 is an explanatory diagram showing functional blocks of an FAQ database server 130 according to the embodiment. [Figure 7] FIG. 10 is an explanatory diagram showing an example of registration of FAQ information according to the embodiment. [Figure 8] 2 is an explanatory diagram showing functional blocks of a degree-of-association calculation server 140 according to the embodiment. FIG. [Figure 9] FIG. 2 is an explanatory diagram showing functional blocks of a user terminal 20 according to the embodiment. [Figure 10] FIG. 10 is a sequence diagram showing an example of the operation of the information processing system according to the embodiment when searching for information. [Figure 11] 10 is an explanatory diagram showing an example of a search screen displayed on the user terminal 20 when accepting an input of a query from a user according to the embodiment. FIG. [Figure 12] 10 is an explanatory diagram showing an example of a search screen displayed on a user terminal 20 when a keyword is displayed in a keyword selection section according to the embodiment. FIG. [Figure 13] 10 is an explanatory diagram showing an example of a search screen displayed on a user terminal 20 when a keyword according to the embodiment is selected. FIG. [Figure 14] FIG. 10 is an explanatory diagram showing an example of a search screen displayed on the user terminal 20 when displaying FAQ candidates according to the embodiment. [Figure 15] 10 is an explanatory diagram showing an example of a search screen displayed on the user terminal 20 when the main text of an FAQ according to the embodiment is displayed. FIG. [Figure 16] 10 is a flowchart of a calculation method for calculating an evaluation value according to the embodiment. [Figure 17] FIG. 10 is an explanatory diagram of a calculation method for calculating an evaluation value according to the embodiment. [Figure 18]FIG. 10 is a schematic diagram illustrating a relevance calculation process according to the embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0021] A preferred embodiment of the present invention will be described below with reference to the accompanying drawings. In this specification and drawings, when it is not necessary to distinguish between components having substantially the same functional configuration or logical significance, the same reference numerals will be used for each of the components, and redundant explanations will be omitted.

[0022] <<1. Overview of the Information Processing System>> An embodiment of the present invention relates to a search system that provides a user with information related to natural language input based on the natural language input by the user. Below, an information processing system according to an embodiment of the present invention will be described using an example of an information processing system that provides FAQs (frequently asked questions) about information devices such as televisions and cameras. FAQs are a collection of past inquiries and operating instructions, and in the embodiment of the present invention, the FAQs are provided to the user as character strings. Note that the embodiment of the present invention is not limited to an information processing system that provides FAQs about information devices, and an information processing system that provides various types of information may also be provided. For example, the information processing system may be a response database used by operators in a call center or a knowledge management system in a company.

[0023] Fig. 1 is an explanatory diagram showing an overview of an information processing system according to an embodiment of the present invention. As shown in Fig. 1, a server group 10 is connected to a user terminal 20 via a network 30. The server group 10 is made up of a web server 110, a keyword evaluation server 120, an FAQ database server 130, and a relevance calculation server 140.

[0024] (1) Server Group 10 The server group 10 is an information processing system that provides information to users. The server group 10 receives character string information from the user terminal 20 via the network 30. Here, the character string information is information generated by the user terminal 20 based on the user's operation on the user terminal 20. The server group 10 analyzes the character string information. The server group 10 extracts keywords from the character string information, presents the extracted keywords to the user, and acquires related information based on the combination of keywords selected by the user. The server group 10 also performs a feedback process in accordance with feedback information that indicates whether the user's problem has been solved, thereby learning to present information to the user in a more appropriate order.

[0025] (2) User terminal 20 The user terminal 20 is a terminal device that a user uses to search for FAQs. By operating the user terminal 20, the user can use the server group 10 to perform an information search.

[0026] Browser software (hereinafter referred to as "browser") for connecting to the server group 10 via the network 30 is installed on the user terminal 20. The user terminal 20 can use an application to send and receive user operation information and related information from the server group 10. The user sets a URL to a web page provided by the server group 10 on the browser, and the user terminal 20 receives information related to the display of a web page for performing a search from the server group 10. The user inputs a search query in a specified field on the web page displayed on the screen. In this embodiment, the search query is a question input in the form of natural language.

[0027] The user terminal 20 transmits the search query to the server group 10. The user selects keywords using the user terminal 20. The user terminal 20 transmits the selected keyword combination to the server group 10 and receives FAQ information related to the combination from the server group 10. By selecting an FAQ using the user terminal 20, the user can obtain information related to the keyword combination from the server group 10 and view it by displaying it on the display unit 23 of the user terminal 20. In the embodiment of the present invention, the user terminal 20 is described as a smartphone, but the user terminal 20 to which the present invention can be applied is not limited to a smartphone and any terminal can be used.

[0028] The information processing system according to the embodiment of the present invention has been outlined above with reference to Fig. 1. Next, the hardware configuration according to the embodiment of the present invention will be described with reference to Fig. 2.

[0029] <<2. Hardware Configuration>> The information processing of the above-described information processing system is realized by cooperation between software and the hardware of each server of the server group 10, which will be described below.

[0030] 2 is a block diagram showing an example of the hardware configuration of the Web server 110, keyword evaluation server 120, FAQ database server 130, and relevance calculation server 140 that constitute the server group 10. Each server that constitutes the server group 10 includes a CPU (Central Processing Unit) 101, a ROM (Read Only Memory) 102, a RAM (Random Access Memory) 103, and a system bus 104. Each server that constitutes the server group 10 also includes an input unit 105, a display unit 106, a storage device unit 107, and a communication interface unit 108.

[0031] The CPU 101 functions as a general-purpose arithmetic processing unit and control device, and controls the operations of each server according to various programs. The CPU 101 may be a microprocessor. The ROM 102 temporarily stores programs and calculation parameters used by the CPU 101. The input unit 105, display unit 106, and storage unit 107 are provided as needed and are used for program input, etc. The input unit 105 is an input device such as a mouse, touch panel, or keyboard. The display unit 106 is a display device such as an LCD display or an OLED display. The storage unit 107 is an internal storage device such as a HDD or SSD, or an external storage device such as an SD card or USB memory. The communication interface unit 108 is an interface that handles communication between terminals and may be a wired connection such as a LAN port or a wireless standard such as Wi-Fi (registered trademark). The above components are interconnected by a system bus 104. Specifically, the system bus 104 may be, but is not limited to, PCIe (registered trademark) or NVM Express (registered trademark). The input unit 105, the display unit 106, the storage device unit 107, and the communication interface unit 108 are controlled by the CPU 101 and software stored in the ROM 102 or the RAM 103. More specifically, a program stored in the ROM 102 or the storage device unit 107 is read into the RAM 103, and the CPU 101 and the RAM 103 work together to function as a control unit.

[0032] The hardware configuration of each terminal according to the embodiment of the present invention has been described above with reference to Fig. 2. Next, an example of the configuration of an information processing system according to the embodiment of the present invention will be described.

[0033] <<3. Example of information processing system configuration>> An example of the configuration of an information processing system according to an embodiment of the present invention will be described with reference to Fig. 1. An example of the configuration of an information processing system according to an embodiment of the present invention will be described below with reference to Fig. 1.

[0034] <3-1. Server group configuration example> The functions of the servers constituting the server group 10 and the information stored in the storage unit will be described in detail with reference to FIGS. 3, 4, 5, 6, 7, 8 and 9.

[0035] (1) Web Server 110 3 is a functional block diagram of the Web server 110 according to this embodiment. The Web server 110 has a function of receiving a query from the user terminal 20 and maintaining a connection between the user terminal 20 and the server group 10 until the user's information search is completed. The Web server 110 also has a function of transmitting information related to various searches received from the user terminal 20 to each device constituting the server group 10 and receiving the processing results. The Web server 110 also has a function of transmitting the processing results of each device in the server group 10 to the user terminal 20 and displaying them on the user terminal 20. Below, the functional units of the Web server 110 will be described in detail with reference to FIG. 3.

[0036] The control unit 111 executes software stored in a storage unit (not shown) to control the functions described below as a web server. The control unit 111 also controls the communication interface unit 114 to realize information communication with other terminals.

[0037] The session management unit 112 receives a connection request from the user terminal 20, identifies the target user terminal 20, and performs processing to enable continuous communication until the user finishes the search. The session management unit 112 accepts a connection from the user terminal 20, issues identification information such as a cookie or token, transmits it to the user terminal 20, and establishes a session with the user terminal 20 by attaching the cookie or token information to subsequent communication information from the user terminal 20. The session management unit 112 maintains the session and identifies the terminal during a series of search processes based on the cookie or token information. This specification omits descriptions of the cookie and token information associated with each transmission and reception, and the processing for identifying the user terminal 20 based on the information. The inclusion of the session management unit 112 enables the Web server 110 to process even multiple user terminals 20.

[0038] The display control unit 113 transmits to the user terminal 20 the content to be displayed on the display unit 23 of the user terminal 20 (described later), and controls the display so that it is displayed on the display unit 23. When it is necessary to update the display unit 23 as a result of processing by each device in the server group 10 or when receiving information from the user terminal 20, the display control unit 113 sends to the user terminal 20 a display control signal that changes the content displayed on the screen. The display control signal includes information related to the display content, such as keyword information for keywords to be displayed or FAQ information for FAQs to be displayed.

[0039] The communication interface unit 114 performs communication processing to send and receive communications with the user terminal 20 and other terminals that make up the server group 10. The communication interface unit 114 is connected to the network in either a wired or wireless manner, and receives instructions from the control unit 111 to communicate with other terminals that make up the server group 10. Alternatively, the communication interface unit 114 receives instructions from the control unit 111 to communicate with the user terminal 20.

[0040] The input receiving unit 115 receives a search query input via the input unit 22 of the user terminal 20. In this embodiment, the search query will be described as being in natural language. The input receiving unit 115 has a function of transmitting the search query to the keyword evaluation server 120. In addition, the input receiving unit 115 receives the content selected by the user via the input unit 22 of the user terminal 20.

[0041] (2) Keyword evaluation server 120 Fig. 4 is a functional block diagram of the keyword evaluation server 120 according to this embodiment. The keyword evaluation server 120 has the functions of extracting keywords from natural sentences received from the Web server 110, calculating evaluation values ​​for the keywords, and transmitting the calculation results to the Web server 110. The keyword evaluation server 120 also has a keyword information storage unit 123 that stores keyword information. The functional units of the keyword evaluation server 120 will be described in detail below with reference to Fig. 4.

[0042] The control unit 121 executes software stored in a storage unit (not shown) to realize control of the functions described below as the overall control of the keyword evaluation server 120. The control unit 121 also controls the communication interface unit 125 to realize information communication with other terminals.

[0043] The keyword extraction unit 122 extracts keywords from the natural text received by the Web server 110. The keyword evaluation server 120 refers to the keyword information storage unit 123 (described later) and performs a character string search to determine whether the natural text contains any keywords.

[0044] The keyword information storage unit 123 stores a keyword information table 50. The keyword information table 50 stores, as keyword information, keywords, synonym information of the keywords, and keyword priority settings in association with each other.

[0045] (Keyword information) Keyword information will be specifically described below with reference to Fig. 5. Fig. 5 is an example of data stored as keyword information in keyword information table 50 of keyword information storage unit 123. As shown in Fig. 5, the data items of keyword information table 50 that make up the keyword information are Keyword 51, which stores keywords as character strings, Synonym 52, which stores synonyms as character strings, and Priority 53, which stores keyword priority settings.

[0046] The keyword 51 is a primary key and does not overlap throughout the keyword information table 50. In the keyword 51, keywords that the administrator of the keyword information storage unit 123 considers important are registered in advance.

[0047] The synonym 52 is synonym information related to the keyword. For example, if "television" is registered as the keyword 51, "television" or the like is set as the synonym 52. Furthermore, multiple synonyms 52 may be set for one keyword 51.

[0048] Priority 53 is a priority setting for a keyword. Priority 53 is usually set to a value of False. Among the items set in Keyword 51, there are some that are considered to be important words included in a question, such as product names and model numbers. The control unit 121 receives an operation from a database administrator or a user and sets Priority 53 to True for such important keywords. The display control unit 113 may control the display on the user terminal 20 to prioritize keywords for which Priority 53 is set to True in the display control described below.

[0049] The keyword information stored in the keyword information storage unit 123 has been described above in detail with reference to FIG.

[0050] 4 calculates an evaluation value based on the keywords extracted by the keyword extraction unit 122. The evaluation value calculation unit 124 calculates an evaluation value for each extracted keyword while referring to the FAQ information stored in the FAQ database server 130, and transmits the evaluation values ​​to the Web server 110 together with the extracted keyword information.

[0051] The communication interface unit 125 transmits and receives communications with other terminals constituting the server group 10. The communication interface unit 125 is connected to the network in either a wired or wireless manner, and receives instructions from the control unit 121 to communicate with other terminals constituting the server group 10.

[0052] (4) FAQ database server 130 6 is a functional block diagram of the FAQ database server 130 according to this embodiment. The FAQ database server 130 has an FAQ information storage unit 132 that stores information related to FAQs (FAQ information) to be provided to users. When it becomes necessary for another terminal constituting the information processing system to refer to FAQ information, the FAQ database server 130 has a function of providing the FAQ information to the other terminal in response to a request from the other terminal.

[0053] The control unit 131 executes software stored in a storage unit (not shown) to realize control of the functions described below as the FAQ database server 130. The control unit 131 also controls the communication interface unit 133 to realize information communication with other terminals.

[0054] The FAQ information storage unit 132 stores an FAQ information table 60. The FAQ information table 60 stores, as FAQ information, an FAQ-specific ID (FAQ_ID 61), an FAQ title (FAQ_Title 62), a FAQ main text (FAQ_Main 63), and a category to which the FAQ belongs (FAQ_Category 64), in association with each other.

[0055] (FAQ information) FAQ information will be described in detail below with reference to Fig. 7. FAQ information is information related to FAQs provided to users. Fig. 7 is an example of FAQ information stored in the FAQ database server 130. As shown in Fig. 7, the data items of the FAQ information table 60 are composed of FAQ_ID 61 that stores an ID unique to the FAQ, FAQ_Title 62 that stores the title of the FAQ that simply indicates the content of the FAQ, FAQ_Main 63 that stores a character string that indicates the FAQ itself, and FAQ_Category 64 that stores the category to which the FAQ belongs.

[0056] The FAQ_ID 61 is a primary key and is a number that is unique throughout the table. The FAQ_ID 61 may be an automatically assigned numeric string. The FAQ_ID 61 is used to search the entire FAQ information table 60 for a saved row of a specific FAQ.

[0057] FAQ_Title 62 is the title of the FAQ provided to the user in the present invention. It corresponds to FAQ_Main 63 information described later, and is a character string that succinctly expresses the content of the FAQ_Main 63 information.

[0058] FAQ_Main 63 is the main text of the FAQ to be provided to the user in the present invention. The main text of the FAQ is information provided to the user expressed as a character string. For example, a procedure for dealing with a TV malfunction is stored as a character string.

[0059] FAQ_Category64 is information indicating which product field the FAQ relates to. More specifically, FAQ_Category64 may store, for example, TV recorders, cameras, audio, and others. Note that the above is just an example, and the administrator of the FAQ database server 130 may decide which FAQ_Category64 to associate with an FAQ, taking into account the technical field to which the FAQ applies.

[0060] The keyword information stored in the FAQ information storage unit 132 has been described above in detail with reference to FIG.

[0061] 6 transmits and receives communications with other terminals constituting the server group 10. The communication interface unit 133 is connected to the network in either a wired or wireless manner, and receives commands from the control unit 131 to communicate with other terminals constituting the server group 10.

[0062] (4) Relevance calculation server 140 8 is a functional block diagram of the relevance calculation server 140 according to this embodiment. The relevance calculation server 140 calculates the relevance based on the combination of keywords selected by the user, which is received by the input receiving unit 115 of the Web server 110, and the FAQ information received from the FAQ database server 130.

[0063] The control unit 141 executes software stored in a storage unit (not shown) to realize control of the function described below as a relevance calculation function. The control unit 141 also controls the communication interface unit 144 to realize information communication with other terminals.

[0064] The relevance calculation unit 142 calculates the relevance based on the combination of keywords selected by the user received by the input receiving unit 115 of the Web server 110 and the FAQ information received from the FAQ database server 130. The relevance calculation unit 142 acquires the combination of keywords selected by the user and the FAQ information from the FAQ database server 130 via the communication interface unit 145, and calculates the relevance. More specifically, the relevance calculation unit 142 has a neural network 1. The neural network 1 receives an FAQ_ID 61 and the combination of keywords selected by the user. The neural network 1 outputs a value of the relevance from the input as a numerical value. The neural network 1 may be configured in the form of a CNN (Convolutional Neural Network) or an RNN (Recurrent Neural Network), or may be a combination of these.

[0065] The relevance is used to determine the FAQ to be displayed in the FAQ candidate display section 204 on the screen 200 of the user terminal 20, which will be described later. The display control section 113 performs display control so that keywords are preferentially displayed in the FAQ candidate display section 204 in descending order of relevance. The neural network 1 receives as input a keyword combination selected by the user and past feedback information, and outputs the relevance. The neural network 1 is a trained model that has learned, through training, the relationship between the keyword combination selected by the user, past feedback information, and the relevance. Note that any machine learning algorithm may be used as the neural network 1. For example, the relevance may be learned and inferred using a method such as a random forest or a support vector machine.

[0066] The feedback unit 143 receives, via the communication interface unit 144, the feedback information input to the input unit 22 from the user terminal 20 that was received by the Web server 110 in S16 described above. The feedback unit 143 performs feedback processing based on the received feedback information. Furthermore, the feedback unit 143 sends a command to the display control unit 113 of the Web server 110 to display, on the display unit 23 of the user terminal 20, a feedback button display unit 206 that asks whether the user was able to find the information they desired. When the feedback information received indicates that the user was unable to find the information they desired, the feedback unit 143 adjusts the calculation parameters so that the relevance calculation server 140 outputs a smaller value of the relevance. When the feedback information received indicates that the user was able to find the information they desired, the feedback unit 143 adjusts the calculation parameters so that the relevance calculation server 140 outputs a larger value of the relevance.

[0067] The communication interface unit 145 transmits and receives communications with other terminals constituting the server group 10. The communication interface unit 145 is connected to the network in either a wired or wireless manner, and receives instructions from the control unit 141 to communicate with other terminals constituting the server group 10.

[0068] The functions of the servers constituting the server group 10 of the information processing system according to the embodiment of the present invention and the information stored in the storage unit have been described above. Next, an example of the functions of the user terminal 20 of the information processing system according to the embodiment of the present invention will be described.

[0069] <3-2. User terminal 20> The user terminal 20 is connected to the server group 10 via the network 30, and performs searches in cooperation with the server group 10 in response to user operations.

[0070] 9 is a functional block diagram of the user terminal 20 according to this embodiment. The control unit 21 executes software stored in a storage unit (not shown) to realize control of the functions described below. The control unit 21 also controls the communication interface unit 24 to realize information communication with other terminals.

[0071] The input unit 22 provides the user terminal 20 with functions for the user to perform input operations. The input unit 22 is composed of input devices such as a touch panel, a keyboard, a mouse, or buttons, and a program that controls them. The input unit 22 accepts input operations from the user, such as inputting a search query, selecting a keyword, and replying to feedback information. The input operations and feedback information in the input unit 22 are transmitted to the input accepting unit 115 of the Web server 110 via the communication interface unit 24.

[0072] The display unit 23 is composed of a display device such as a liquid crystal display or an OLED display, and a program for controlling it. The display unit 23 displays according to a signal related to screen display received from the display control unit 113 of the Web server 110 via the communication interface unit 24.

[0073] The communication interface unit 24 transmits and receives communications with the terminals of the server group 10. The communication interface unit 24 is connected to the network 30 in either a wired or wireless manner, and receives instructions from the control unit 21 to communicate with the terminals that make up the server group 10.

[0074] A more detailed description will be given with reference to FIGS. 11, 12, 13, 14, and 15, which will be described later. Screen 200 is displayed on display unit 23 and presents information related to the search process to the user. Screen 200 has an input field 201 for inputting a subject to be searched for, and the user inputs a search query 202 into input field 201. Screen 200 also has a keyword selection section 203, and the user can select a keyword via input unit 22 of user terminal 20. Screen 200 also has an FAQ candidate display section 204 that displays candidates for FAQs to be presented, and the user can select the title of any FAQ via input unit 22. Screen 200 also has an FAQ answer display section 205, and displays the FAQ selected in FAQ candidate display section 204. Screen 200 also has a feedback button display section 206, and the user can select via input unit 22 of user terminal 20 whether the FAQ information displayed in FAQ answer display section 205 was useful.

[0075] An example of the functions of the user terminal 20 of the information processing system according to the embodiment of the present invention has been described above. Next, the network configuration of the information processing system according to the embodiment of the present invention will be described.

[0076] <3-3. Network configuration example> 1, each server in the server group 10 is connected to a local network and connected to a network 30 via a router (not shown). In the embodiment, the network 30 is described as the Internet, but is not limited to this and may be a local network. Furthermore, each server in the server group 10 does not have to be located in the same segment, and may communicate via the Internet, for example, and function as the server group 10.

[0077] <<4. Example of operation>> Below, an example of operation of the information processing system of an embodiment of the present invention will be explained with reference to Figure 10, which is a sequence diagram of the search process, and Figures 11, 12, 13, 14, and 15, which are example screens of the user terminal 20.

[0078] Regarding matters related to user session management, methods using cookies and tokens are widely accepted and disclosed, so descriptions of these will be omitted in the description of this embodiment. Also, since widely known configurations such as Ajax are used for processing related to screen control of the user terminal 20, descriptions of these will also be omitted.

[0079] FIG. 11 shows an example of the configuration of a screen 200 on the user terminal 20. The user inputs a URL for accessing a web page that provides a search service on the browser of the user terminal 20. The user terminal 20 transmits a signal to the web server 110 to establish a connection with the server group 10 (S1). In response to the HTTP access from the user terminal 20, the web server 110 transmits an HTML file representing a web page for providing the search service to the user terminal 20 (S2). The user terminal 20 displays a screen 200 including a search form represented by the HTML file on the display unit 23. The user inputs the content to be searched into an input field 201 via the input unit 22. FIG. 11 shows an example of a screen displayed on the screen 200. The search query 202 input into the input field 201 is transmitted from the user terminal 20 to the web server 110 of the server group 10 via the network 30 (S3). The transmission of the search query 202 from the user terminal 20 to the web server 110 is performed in response to input of the enter key or the confirm key.

[0080] The control unit 111 transmits the natural sentence received from the user terminal 20 as a search query to the keyword evaluation server 120 (S4). The keyword extraction unit 122 performs keyword extraction processing (S5) on the natural sentence received from the user terminal 20. The keyword extraction processing extracts pre-registered keywords from the natural sentence. Details of the keyword extraction processing will be described later. The control unit 121 transmits an FAQ information acquisition request (not shown) to the FAQ database server 130. The FAQ database server 130 refers to the FAQ information table 50 in the FAQ information storage unit 132 based on the FAQ information acquisition request, and transmits the FAQ information to the keyword evaluation server 120. In this way, the evaluation value calculation unit 124 acquires the FAQ information from the FAQ information storage unit 132 of the FAQ database server 130 (S6). The keyword evaluation server 120 performs evaluation value calculation processing (S7) based on the group of keywords that are the processing results of the keyword extraction processing, the Priority 53 of those keywords, and the FAQ information acquired in S6. Details of the evaluation value calculation processing will be described later.

[0081] The keyword evaluation server 120 transmits the keywords and evaluation values, which are the processing results of the evaluation value calculation process, to the Web server 110 (S8). The display control unit 113 of the Web server 110 transmits display control information to the user terminal 20 so as to display the keywords in descending order of evaluation value (S9). The display control unit 113 performs display control so as to display the keywords in the keyword selection unit 203 of the screen 200 in descending order of evaluation value. FIG. 12 shows an example of the screen 200 in which keywords are displayed in the keyword selection unit 203.

[0082] The input unit 22 accepts a selection operation of a keyword that the user considers appropriate for the search query from among the keywords displayed in the keyword selection unit 203 on the screen 200. The input unit 22 transmits the keyword selected by the user to the Web server 110 (S10). The keyword selected by the user is displayed in a highlighted manner. FIG. 13 is an example of the screen 200 on which the selected keyword is displayed in a highlighted manner. FIG. 13 shows an example in which two keywords, "television" and "malfunction", have been selected. The input acceptance unit 115 transmits the selected keyword information to the relevance calculation server 140 via the communication interface unit 114 (S11). The relevance calculation unit 142 acquires FAQ information stored in the FAQ information storage unit 132 of the FAQ database server 130 (S12).

[0083] The relevance calculation unit 142 performs relevance calculation processing based on the selected keyword and FAQ information (S13). The relevance calculation processing (S13) will be described in detail later. The relevance calculation unit 142 transmits the relevance and FAQ information to the Web server 110 (S14).

[0084] The display control unit 113 performs display control so that a character string indicating the title of the FAQ included in FAQ_Title62 of the FAQ information received in S14 is displayed in the FAQ candidate display unit 204 on the screen 200 of the user terminal 20. FIG. 14 is an example of the screen 200 displaying the titles of the FAQ in the FAQ candidate display unit 204. The display control unit 113 controls the display of the FAQs in a list format in descending order of relevance. From the presented FAQ titles, the user selects the FAQ that the user considers to be closest to the content of their search using the input unit 22. The input unit 22 transmits the title of the selected FAQ to the input receiving unit 115. The input receiving unit 115 receives the title of the FAQ and identifies the FAQ_ID61 associated with FAQ_Title62. The control unit 111 transmits the identified FAQ_ID61 to the FAQ database server 130, and also acquires the text of the FAQ linked to FAQ_ID61 transmitted by the FAQ database server 130. The display control unit 113 performs display control so that the text of the acquired FAQ is displayed in the FAQ answer display unit 205 (S15). FIG.

[0085] The user checks the text of the FAQ displayed in the FAQ answer display section 205 of the display section 23, and selects an answer displayed in the feedback button display section 206 to see if it is an answer to their question or a solution to their problem. The selection is made using the input section 22. The input section 22 transmits the selected information to the Web server 110 as feedback information (S16). The input acceptance section 115 of the Web server 110 receives the selected information. The input acceptance section 115 transmits the feedback information, together with the FAQ information selected by the user, to the feedback section 143 of the relevance calculation server 140 (S17).

[0086] The feedback section 143 performs a feedback process (S18) on the relevance calculation section 142. The feedback process will be described in detail later.

[0087] The data exchange and data processing outline and screen examples related to this embodiment have been described above using the sequence diagram in Fig. 10 and the explanatory diagrams of screen 200 in Fig. 11, Fig. 12, Fig. 13, Fig. 14, and Fig. 15. Next, each data processing step (S5, S7, S13, S18) will be described in detail.

[0088] (S5: Keyword extraction process) Keyword extraction processing involves extracting keywords related to words contained in a search query. Keyword extraction processing is divided into two steps: (1) a step of performing morphological analysis of natural language, and (2) a step (keyword inclusion determination) of determining whether keywords are contained in the morphemes contained in the natural language (the analysis results of (1)). Each step is explained in detail below.

[0089] (1) Morphological analysis of natural language The keyword extraction unit 122 divides the natural sentence received from the user terminal 20 into morphemes. For each morpheme divided by morphemes, the keyword extraction unit 122 excludes morphemes that are not considered to affect the search based on the part of speech of each morpheme, and sets the result as the analysis result of (1). For example, conjunctions, interjections, or auxiliary verbs are excluded.

[0090] (2) Keyword content judgment The keyword inclusion determination process is performed with reference to the keyword information storage unit 123. The keyword extraction unit 122 acquires the Keyword 51, Synonym 52, and Priority 53 of a row identified by one keyword from the keyword information table 50 stored in the keyword information storage unit 123. The keyword evaluation server 120 determines by character string comparison whether the word resulting from the analysis in (1) matches the word in Keyword 51 or a word registered in Synonym 52. If the result of the character string comparison shows a match with Keyword 51 or a word registered in Synonym 52, the evaluation value calculation unit 124 determines that the keyword set in Keyword 51 is a keyword included in natural language text.

[0091] The keyword extraction unit 122 performs the above process on all keywords stored in the keyword information storage unit 123, and sets the keywords contained in the natural sentence as the processing result in S5. The keyword extraction process has been described above in detail.

[0092] (S7: Evaluation value calculation process) The method for calculating the evaluation value of each keyword will be described in detail below with reference to Figures 16, 17, and 18. Figure 16 is a flowchart showing the processing of S7. In the following explanation, the calculation of evaluation values ​​will be described assuming that the keywords "expense," "television," and "malfunction" were extracted in S5. At each stage of the processing, the calculation results are stored in a memory unit (not shown) provided in evaluation value calculation unit 124.

[0093] The evaluation value calculation unit 124 calculates an evaluation value for the FAQ based on the keyword information acquired in S5 and the FAQ information acquired in S6. The evaluation value calculation unit 124 determines that the above two pieces of information have been acquired and starts the evaluation value calculation process (S701). For each category identified by FAQ_Category 64, the evaluation value calculation unit 124 calculates the total number of FAQs classified into that category (S702). In FIG. 17, if there are 1,000 FAQs classified into the category "TV / Recorder," 1,000 is stored in the storage area 602 in the storage unit. The same applies to other categories.

[0094] The evaluation value calculation unit 124 calculates the number of FAQs containing the keyword "television" within a category based on the FAQ_Main63 information and category for the FAQ identified by each FAQ_ID. Specifically, the evaluation value calculation unit 124 calculates the evaluation value as follows: The Web server 110 has a storage unit that stores, for each keyword, the number of FAQs containing that keyword within each category, and the initial value is 0. The evaluation value calculation unit 124 performs a string search to determine whether the keyword "television" exists in the text of the FAQ represented by the FAQ_Main63 information for all FAQ_IDs (S704). If a certain keyword exists, the evaluation value calculation unit 124 adds 1 to the storage area of ​​the storage unit for the category corresponding to the FAQ information (S705). For example, if the keyword "television" is included in an FAQ in the category "television / recorder," 1 is added to the storage area 601 in FIG. 17.

[0095] The evaluation value calculation unit 124 calculates the total number of FAQs in the category and the ratio of the number of FAQs in FAQ_Main that contain the keyword "television" in the text for each category, and sets the sum of the ratios calculated for each category as the evaluation value for that keyword (S706). The evaluation value calculation unit 124 saves the relevance and keyword combination in a storage unit (not shown) and ends the evaluation value calculation process (S707). The information saved in the storage unit (not shown) is sent to the Web server 110 in S8 of FIG. 10.

[0096] By using the above-described calculation method for the evaluation value, the search system can present to the user keywords related to information in a category that has only a small number of FAQs. For example, if there is only one FAQ in the category "Audio" (not shown), a method of calculating the relevance based on the total number of FAQs in the entire category would not be able to distinguish and present to the user keywords related to that one FAQ. By using the evaluation value calculation method of this embodiment, the display control unit 113 can display to the user FAQs even if they belong to a category that has only a small number of FAQs.

[0097] Furthermore, the evaluation value calculation unit 124 calculates the number of FAQs for each category that contain a keyword by determining whether the keyword matches the word, but the calculation may also take into consideration words associated with the keyword as appropriate. More specifically, the evaluation value calculation unit 124 may refer to Synonym 52 in the keyword information table 50 and determine whether synonyms are included in the main text of the FAQ.

[0098] Furthermore, when Priority53 included in the keyword information is set to True, the evaluation value calculation unit 124 may add a fixed value to the evaluation value of the keyword instead of always displaying it. Also, the value to be added to the evaluation value of the keyword may be stored as a numerical value in the Priority53 field.

[0099] The evaluation value calculation process has been explained in detail above. By using the above calculation method for the evaluation value, the search system can present to the user information even for categories with only a small number of FAQs.

[0100] (S13: Relevance calculation process) FIG. 18 is a diagram illustrating the relevance calculation process in S13. In S13, the relevance calculation unit 142 calculates the relevance between the keywords selected by the user and all FAQs using a combination of keywords selected by the user. The relevance calculation unit 142 receives the FAQ_ID of the FAQ and one or more keywords as input, and outputs the relevance (numerical value). More specifically, the relevance calculation unit 142 converts the input keywords into word vectors using a technique such as Word2Vec, calculates the sum of the word vectors of the keywords selected by the user, and inputs the sum to the neural network 1. The neural network 1 performs a calculation process on the sum of the input word vectors using the weights of each node constituting the network, and transmits the result to subsequent nodes to calculate the relevance. Here, the relevance is the output of the neural network 1, which has learned the number of keywords included in the main text of the FAQ in the process described below. In this embodiment, the initial value of the weight of each node in the neural network 1 is 0. The weight value is updated as the learning of the neural network 1 progresses.

[0101] The neural network 1 is trained by a method in which learning proceeds through feedback processing, which will be described later, during actual use. Therefore, before learning through feedback processing has been performed, the relevance calculation unit 142 may not be able to calculate an appropriate relevance. For example, if relevance learning has not been performed for a combination of a certain FAQ_ID 61 and a keyword, the weight of each node is 0, and therefore the relevance for that combination is output as 0. Therefore, the relevance calculation unit 142 replaces the relevance with a value calculated by another means, depending on whether feedback processing has already been performed for the input FAQ_ID 61. More specifically, if the number of feedback processes for the combination of FAQ_ID 61 and keyword for which relevance calculation is to be performed is 0, the relevance calculation unit 142 replaces the relevance of the FAQ with the number of keywords included in the text of the FAQ corresponding to the FAQ_ID. The combination of FAQ_ID 61 of the FAQ that has undergone feedback processing and the selected keyword is stored in a storage unit (not shown). By referring to the combination of FAQ_ID61 of this FAQ and a keyword, it is possible to determine whether feedback calculation processing has been performed in the past for the combination of FAQ_ID61 and a keyword for which relevance calculation processing is currently being performed.

[0102] In the above example, the neural network 1 has been described as having an initial weight value of 0 for each node, but a neural network 1 that has undergone pre-training may be used as appropriate. For example, the pre-training uses FAQ_ID 61 and the keywords and their numbers contained in the text of the FAQ. In this case, the keywords and their numbers contained in the text of the FAQ are acquired and calculated by referring to keyword information table 50.

[0103] By repeatedly performing this relevance calculation process and the feedback process described below, the relevance calculation unit 142 can present, for a huge number of keyword combinations, FAQs that contain the keywords, as well as FAQs that do not contain the keywords but are relevant.

[0104] (S18: Feedback processing) In S18, the feedback unit 143 changes the parameters of the neural network 1 used in S13 in the next and subsequent search operations in accordance with the feedback information from the user. More specifically, in S18, the feedback unit 143 adjusts the calculation parameters (weights) of each node constituting the neural network 1 using the backpropagation method for the neural network of the relevance calculation unit 142.

[0105] The feedback unit 143 determines the magnitude of the error value to be backpropagated as either the maximum or minimum value of the output of the neural network 1. Alternatively, it may be determined based on the number of keywords selected by the user that were included in the FAQ. For example, even if the user selects feedback information of "No," backpropagation of the error does not need to be performed if all keywords are included in the FAQ. By adopting this configuration, the feedback unit 143 can reduce the impact even if feedback processing is performed on incorrect feedback information due to the special circumstances of each user. As mentioned above, in the case of the first feedback for the combination of the FAQ_ID and keyword, the number of keywords included in the body of the FAQ is unconditionally learned as the relevance.

[0106] <<5. Effects>> The Web server 110 of the server group 10 presents multiple keywords to the user via the user terminal 20, and displays information based on the combination of keywords selected by the user, allowing the user to quickly search for information that matches their purpose. In addition, because the user explicitly selects multiple keywords, the search results are more convincing.

[0107] <<6. Modifications>> Modifications of the embodiments of the present invention will be described below. Note that each of the modifications described below may be applied to the embodiments of the present invention alone within the scope of the present invention. Also, they may be applied to the embodiments of the present invention in a combined manner. Furthermore, the modifications may be applied in place of the configurations described in the embodiments of the present invention, or may be configured in addition to them.

[0108] (Variation 1) In the above embodiment, an example was described in which the search query 202 is input to the input unit 22 as natural language. However, the search query 202 may be speech acquired from the input unit 22 of the user terminal 20 via a speech input unit (not shown). In this modification, the input unit 22 of the user terminal 20 has a microphone for inputting speech in addition to the configuration of the embodiment. The input receiving unit 115 of the web server 110 has a conversion function for converting the speech into text. Furthermore, after receiving the search query in S3, speech conversion processing may be performed before S4 is performed. By adopting the above configuration, the user can perform a search even when the input is speech.

[0109] (Variation 2) In the above embodiment, the evaluation value is calculated in S7 for each search. However, the evaluation value may be calculated in advance. Since the evaluation value changes when new information or keywords are added to the search target, the evaluation value calculation unit 124 calculates an evaluation value for each keyword when new information or keywords are added to the search target, and the keyword information table 50 may store the evaluation value in advance as an added item. Specifically, when a keyword is added to or deleted from the keyword information table 50 or when keyword information is changed, or when an FAQ is added to or deleted from the FAQ information table 60 or when FAQ information is changed, the control unit 111 performs an evaluation value calculation process for the keyword. The control unit 111 stores the calculated processing result in a memory unit (not shown) in association with the keyword. In S7 in the above embodiment, the evaluation value calculation unit 124 obtains the evaluation value for the keyword by referencing the memory unit. The above configuration eliminates the need to calculate an evaluation value for each search, thereby reducing the load on the search system.

[0110] (Variation 3) In the above embodiment, a configuration has been described in which all keywords obtained as a result of keyword extraction are displayed. However, the keywords to be displayed may be determined as appropriate based on the contents of the user terminal or the system to be applied. More specifically, the Web server 110 may perform display control so that keywords for which Priority 53 is set to True are displayed regardless of their evaluation value, assuming that they have been prioritized. Furthermore, when there are many keywords, the display control may be performed so that only those with high evaluation values ​​are displayed. Furthermore, the number of keywords to be displayed may be adjusted as appropriate depending on the number of keywords for which display control is performed in S7, the size of the display area of ​​the screen 200 of the user terminal 20, and the type of the user terminal 20.

[0111] Furthermore, the display control unit 113 may preselect keywords having evaluation values ​​equal to or greater than a threshold or keywords having Priority 53 set to True on the user terminal 20. More specifically, the display control unit 113 may instruct the user terminal 20 to highlight and display the keywords having evaluation values ​​equal to or greater than a threshold or keywords having Priority 53 set to True. At this time, the keywords may be displayed in a different color, or the like, to distinguish them from the keywords selected by the user. Here, the highlighting is not limited to changing the color. Keywords may be highlighted by methods such as enlarging the text, italics, and underlining. Keywords may also be highlighted by a combination of these methods. The keywords selected in this way are transmitted to the relevance calculation server 140 together with the keywords selected by the user in S11, unless the user cancels the selection state via the input unit 22. The keywords are then treated as keywords for obtaining related FAQ information together with the keywords selected by the user.

[0112] By adopting the above-described modification, even if the search query entered by the user contains a large number of keywords, it is possible to provide the user with a simple search method, thereby further reducing the time and effort required for the user to perform a search. In addition, by selecting important keywords in advance, the user does not need to perform the operation of selecting important keywords.

[0113] (Variation 4) In the above-described embodiment, the relevance calculation is performed using the FAQ_ID61 of the FAQ and the word vector of the keyword selected by the user as input. However, the neural network 1 may be replaced by other input formats and calculation methods. For example, the neural network 1 may use, as input, numerical information obtained by converting the text information of the FAQ into a vector representation. In this case, the relevance calculation unit 142 converts the text information of the FAQ into a vector representation of FAQ_Main63 linked to FAQ_ID61. The FAQ database server 130 additionally has a function of vectorizing the text information of the FAQ included in the FAQ information acquired by the relevance calculation server 140 in S12 using a method such as Seq2vec.

[0114] In this case, in the above-described embodiment, the relevance calculation unit 142 calculated whether or not a feedback process had been performed in the past by having a storage unit (not shown) store a combination of the FAQ_ID61 of the FAQ and a keyword, but in this modified example, the relevance calculation unit 142 determines whether or not a feedback process had been performed in the past by referring to the vector representation of the text information of the FAQ that was previously feedback processed. More specifically, the relevance calculation unit 142 may calculate the cosine similarity with the vector representation of the text information of the past FAQ, and determine whether or not to perform initial processing on the relevance depending on the magnitude of the cosine similarity.

[0115] Alternatively, the neural network 1 may be configured as a network that receives a combination of keywords as input and outputs a plurality of FAQ_IDs 61 that are determined to have a high degree of relevance. In this case, the relevance calculation unit 142 acquires the FAQ_ID 61, which is the output of the neural network 1, based on the keywords selected by the user in the relevance calculation process of S13. The relevance calculation unit 142 transmits the FAQ_ID 61 to the Web server 110 (a process that replaces S14). The Web server 110 performs the process of S15 based on the FAQ_ID 61.

[0116] Furthermore, the method for calculating the relevance is not limited to a method using a neural network. For example, the relevance may be determined based on the results of statistical processing of the distribution of similarities between words appearing in the text of the FAQ and keywords selected by the user. Specifically, in evaluating the similarity between words appearing in the text of the FAQ and keywords selected by the user, the relevance calculation unit 142 may obtain the similarity by calculating the cosine similarity of word vectors converted by Word2Vec.

[0117] By adopting the above-described modified example, the relevance calculation unit 142 can perform information search for an FAQ newly stored in the FAQ information storage unit 132 by making use of feedback information for other FAQs.

[0118] The above describes the modified examples according to the embodiment of the present invention.

[0119] <<7. Supplementary Information>> Although the preferred embodiments of the present invention have been described in detail above with reference to the accompanying drawings, the present invention is not limited to these examples. It is clear that those skilled in the art can easily create various modifications within the scope of the technical ideas described in the claims, and these modifications also naturally fall within the technical scope of the present invention.

[0120] Note that the steps in the processing of the server group 10, each server constituting the server group 10, and the user terminal 20 in this specification do not necessarily have to be processed in the order described in the flowcharts. As long as the effects of the present invention are achieved, the steps in the processing of the server group 10 and the user terminal 20 in this specification may be processed in an order different from the order described in the flowcharts. Furthermore, processing may be performed in parallel, additional processing steps may be included, or some processing steps may be omitted.

[0121] Furthermore, each server of the server group 10 and the user terminal 20 described herein may be implemented as a separate device. Alternatively, some or all of them may be implemented as separate devices. For example, as described in the embodiment, each function may be implemented as a separate device. For example, the functions of the server group 10 may be implemented as a separate device, or only the user terminal 20 may be implemented as a separate device, or the entire information processing system described in FIG. 1 may be implemented as a separate device. For example, some or all of the functions constituting the server group 10 may be implemented on a cloud server. For example, the server group 10 and the user terminal 20 do not necessarily need to be connected by the network 30. The connection method can be replaced by other methods that a person skilled in the art can devise within the scope of known technology, as appropriate. A configuration not including such a network 30 is also within the technical scope of the present invention.

[0122] It is also possible to create a computer program that causes the hardware, such as the CPU, ROM, and RAM, of each server in the server group 10 and the user terminal 20 to perform functions equivalent to those described in the embodiment and each modified example. A storage medium storing the computer program is also provided. [Explanation of symbols]

[0123] 10 Servers 110 Web Server 120 Keyword Evaluation Server 130 FAQ database server 140 Relevance calculation server 20 User terminal 30 Network

Claims

1. An information processing system for searching registered information, an input receiving unit that receives a search query; a keyword extraction unit that extracts keywords related to words included in the search query; an evaluation value calculation unit that calculates an evaluation value of each keyword extracted by the keyword extraction unit based on the proportion of information that includes a word associated with the keyword among the information; a display control unit that performs display control to display the related keywords on the terminal in an order based on the evaluation value; a search unit that searches for information related to a selected keyword among the keywords whose display is controlled by the display control unit, the display control unit performs display control to display information related to the selected keyword on the terminal; The evaluation value calculation unit calculates the ratio of the information containing the word associated with the selected keyword for each category to which the information belongs, and determines the sum of the ratios for each category as the evaluation value. Information processing system.

2. An information processing system for searching registered information, an input receiving unit that receives a search query; a keyword extraction unit that extracts keywords related to words included in the search query; an evaluation value calculation unit that calculates an evaluation value of each keyword extracted by the keyword extraction unit based on the proportion of information that includes a word associated with the keyword among the information; a display control unit that performs display control to display the related keywords on the terminal in an order based on the evaluation value; a search unit that searches for information related to a selected keyword among the keywords whose display is controlled by the display control unit, the display control unit performs display control to display information related to the selected keyword on the terminal; the evaluation value calculation unit calculates an evaluation value of a keyword registered in the information processing system when a keyword is registered, deleted, or changed in the information processing system, or when information is registered, deleted, or changed, and sets the calculated evaluation value as the evaluation value of each keyword. Information processing system.

3. The information processing system according to claim 1 or 2, characterized in that the display control unit performs display control so as to highlight keywords whose evaluation value is equal to or greater than a threshold value or keywords that have been set as a priority.

4. An information processing system for searching registered information, an input receiving unit that receives a search query; a keyword extraction unit that extracts keywords related to words included in the search query; an evaluation value calculation unit that calculates an evaluation value of each keyword extracted by the keyword extraction unit based on the proportion of information that includes a word associated with the keyword among the information; a display control unit that performs display control to display the related keywords on the terminal in an order based on the evaluation value; a search unit that searches for information related to a selected keyword among the keywords whose display is controlled by the display control unit, the display control unit performs display control to display information related to the selected keyword on the terminal, and performs display control to highlight and display keywords having evaluation values ​​equal to or greater than a threshold value or keywords that have been prioritized; the search unit searches for information including, in addition to the selected keyword, a keyword whose evaluation value is equal to or greater than a threshold value or a keyword that has been prioritized. Information processing system.

5. a relevance calculation unit that calculates a relevance between the selected keyword and the information searched by the search unit, The information processing system according to claim 1 , wherein the display control unit determines an order in which the information is displayed based on the relevance.

6. An information processing system for searching registered information, an input receiving unit that receives a search query; a keyword extraction unit that extracts keywords related to words included in the search query; an evaluation value calculation unit that calculates an evaluation value of each keyword extracted by the keyword extraction unit based on the proportion of information that includes a word associated with the keyword among the information; a display control unit that performs display control to display the related keywords on the terminal in an order based on the evaluation value; a search unit that searches for information related to a selected keyword among the keywords whose display is controlled by the display control unit; a relevance calculation unit that calculates the relevance between the selected keyword and the information searched by the search unit; a feedback receiving unit that receives a response as to whether the information whose display has been controlled by the display control unit satisfies the request of the search query, the display control unit determines an order in which the information is to be displayed based on the relevance, and performs display control to display the information related to the selected keyword on the terminal; The relevance calculation unit calculates the relevance based on past responses in the feedback receiving unit. Information processing system.

7. The information processing system according to claim 6, characterized in that the relevance calculation unit calculates the relevance using a model that has been learned using machine learning to determine the relationship between the keyword and the past answers and the relevance.

8. The information processing system according to claim 1 , wherein the input receiving unit receives the search query by voice input.

9. The information processing system according to claim 1 , wherein the registered information is in natural language in which questions and answers are associated with each other.

10. The computer that searches for registered information is an input receiving unit that receives a search query; a keyword extraction unit that extracts keywords related to words included in the search query; an evaluation value calculation unit that calculates an evaluation value of each keyword extracted by the keyword extraction unit based on the proportion of information that includes a word associated with the keyword among the information; a display control unit that performs display control to display the related keywords on the terminal in an order based on the evaluation value; The display control unit controls the display of the information related to the selected keyword. the display control unit performs display control to display information related to the selected keyword on the terminal; the evaluation value calculation unit calculates a ratio of the information including a word associated with the selected keyword for each category to which the information belongs, and calculates the sum of the ratios for each category as the evaluation value; program.

11. An information processing method for an information processing system that searches for registered information, comprising: receiving a search query input to a terminal by an input receiving unit; a step of a keyword extraction unit extracting keywords related to words included in the received search query; an evaluation value calculation unit calculating an evaluation value of each keyword extracted by the keyword extraction unit based on a ratio of the information that includes a word associated with the keyword; a display control step of performing display control so as to display the related keywords on the terminal in an order based on the evaluation value; a step of searching, by a search unit, for the information related to the selected keyword among the keywords displayed on the terminal by the display control; a step of performing display control by a display control unit to display information related to the selected keyword on the terminal; Including, the evaluation value calculation unit calculates a ratio of the information that includes a word associated with the selected keyword for each category to which the information belongs, and sets the sum of the ratios for each category as the evaluation value; An information processing method in which processing is performed by a computer in an information processing system.

Images