JP2011076360A - Apparatus and method for processing information - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an information processing apparatus and an information processing method for controlling the sequence of questions which can narrow down the optimal retrieval objects according to the selection history of each user. <P>SOLUTION: The information processing apparatus 10 includes: a dialog scenario storage part 13 for storing question sentence data to be provided to a user; a dialog history data storage part 14 for storing dialog history data as the answer history of each user to question sentences; a dialog control part 16 for selecting the question sentences to be provided to the user based on the number of items of the dialog history data and the number of items of retrieval object data, and for narrowing down retrieval object data based on the answers of the user to the question sentences; and an input/output control part 11 for executing the provision of the question sentences to the user, the obtaining of the answers to the question sentences from the user, and the provision of the retrieval object data narrowed down based on the answers to the user. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to an information processing apparatus and an information processing method.

  In recent years, various types of information have been databased and provided in large quantities. We can easily obtain the desired information by searching for such information. For example, examples of such information include restaurant information, job information, property information, financial product information, travel information, hotel information, and the like.

  In order to efficiently obtain desired information from such a large amount of information, setting of search conditions is particularly important. As one method for setting search conditions, there is provided a method of narrowing down search objects in stages by a user interactively answering questions provided by the system. In such an interactive search system, the efficiency of arriving at desired information depends greatly on the order in which questions are provided. For example, Patent Document 1 discloses an interactive system that determines the order in which questions are provided based on entropy calculated from the number of search results for each answer option for a question provided by the system.

Japanese Patent Laid-Open No. 2003-255992

  However, in such a system that simply determines the order of the questions based on the number of search results, even if there is a bias in the answers to the questions for each user, the same question is provided to all users However, there is a problem that the search target is not necessarily efficiently narrowed down. For example, in the case of restaurant search, a certain user always goes out by car, and may search only a restaurant with a parking lot every time. Further, for example, since a certain user does not own a car, the presence / absence of a parking lot is not important and may be searched without being specified each time. However, such conditions for each user have not been considered.

  Therefore, the present invention has been made in view of the above problems, and an object of the present invention is to determine the order of questions that can narrow down the optimal search target according to the selection history for each user. It is an object to provide a new and improved information processing apparatus and information processing method that can be controlled.

  In order to solve the above problems, according to an aspect of the present invention, a dialogue scenario storage unit that stores question sentence data to be provided to a user, and dialogue history data that is an answer history for each user with respect to the question sentence Based on the dialogue history data storage unit to be stored, the number of dialogue history data and the number of search target data, a question sentence to be provided to the user is selected, and the search target is based on the user's answer to the question sentence. Dialog control unit for narrowing down data, provision of the question sentence to the user, acquisition of an answer to the question sentence from the user, and provision of the search target data narrowed based on the answer to the user An information processing apparatus having an input / output control unit for controlling execution is provided.

  According to this configuration, the dialogue control unit selects a question sentence to be provided to the user based on the selection history for each user. For this reason, the dialogue control unit can provide the question sentences in the order customized for each user, and the search efficiency can be improved.

  Further, the dialogue control unit may narrow down by the question sentence based on the number of search target data corresponding to each answer of the question sentence and the number of dialogue history data corresponding to each choice. An expected value of the number of search target data that can be calculated may be calculated, and a question sentence to be provided to the user may be selected from a plurality of the question sentences based on the expected value.

  In addition, the dialogue control unit may repeatedly select the question sentence until the number of pieces of the search target data provided to the user becomes equal to or less than a predetermined result number threshold.

  The dialogue history data may include data in a case where an answer to the question sentence is not designated.

  In addition, when the number of the conversation history data corresponding to the user is equal to or less than a predetermined history number lower limit threshold, the conversation control unit uses the number of the conversation history data not limited to the user to inform the user. The question text to be provided may be selected.

  In addition, when the number of the conversation history data corresponding to the user is greater than or equal to a predetermined history number upper limit threshold, the dialogue control unit further narrows down using at least one of time, place, date, and day of the week. However, the question sentence provided to the user may be selected using the number of conversation history data.

  In order to solve the above problem, according to another aspect of the present invention, a dialogue scenario storage unit that stores question sentence data provided to a user, and a dialogue that is an answer history for each user with respect to the question sentence A dialogue history data storage unit that stores history data, a dialogue control unit that selects a question sentence to be provided to the user and narrows down search target data based on the user's answer to the question sentence, and a user of the question sentence An input / output unit that provides the user, provides an answer to the question sentence from the user, and provides the search target data narrowed down based on the answer to the user. An information processing method including a step in which the dialogue control unit selects a question sentence to be provided to the user based on the number of pieces of dialogue history data and the number of pieces of search target data. There is provided.

  As described above, according to the present invention, it is possible to control the order of questions that can narrow down search targets according to the selection history for each user.

It is a mimetic diagram showing an outline of an information processing system concerning one embodiment of the present invention. It is explanatory drawing which shows an example of the screen display of a user apparatus. It is explanatory drawing which shows another example of the screen display of a user apparatus. It is a block diagram which shows the function structure of information processing apparatus. It is explanatory drawing which shows an example of dialog scenario data. It is explanatory drawing which shows an example of question sentence data. It is explanatory drawing which shows an example of choice data. It is explanatory drawing which shows an example of dialog log | history data. It is explanatory drawing which shows an example of search object data. It is a flowchart which shows the whole flow of dialog control. It is a flowchart which shows the flow of a dialog initial process. It is a flowchart which shows the flow of a dialogue scenario data acquisition process. It is a flowchart which shows the flow of a dialogue process. It is a flowchart which shows the flow of a question sentence data acquisition process. It is a flowchart which shows the flow of an option data acquisition process. It is a flowchart which shows the flow of object data search processing. It is a flowchart which shows the flow of an order control process. It is a flowchart which shows the flow of a dialog history data count process. It is a flowchart which shows the flow of object data count processing.

  Exemplary embodiments of the present invention will be described below in detail with reference to the accompanying drawings. In addition, in this specification and drawing, about the component which has the substantially same function structure, duplication description is abbreviate | omitted by attaching | subjecting the same code | symbol.

<Outline of Information Processing System 100>
First, an outline of an information processing system 100 according to an embodiment of the present invention will be described with reference to FIG. FIG. 1 is a schematic diagram showing an outline of an information processing system according to an embodiment of the present invention. Referring to FIG. 1, the information processing system 100 mainly includes an information processing device 10, a user device 20, a base station 30, and a communication network 40.

  The information processing apparatus 10 has a function of an application server that executes an application provided to the user apparatus 20. In the present embodiment, the information processing apparatus 10 provides a user with a restaurant search function.

  The user device 20 is a terminal device that can be connected to the information processing device 10 via a network when operated by a user. The user device 20 can display a screen generated by the information processing apparatus 10 on the display unit using a service provided by the information processing apparatus 10. The user device 20 is a mobile phone, for example. Further, the user device 20 may be a terminal device such as a PC (Personal Computer) or a PDA (Personal Digital Assistant). In the present embodiment, the user device 20 is connected to the information processing device 10 via the base station 30 and the communication network 40.

  The base station 30 is a terminal device of the mobile phone network. The base station 30 has a function of performing wireless communication with the user device 20 when the user device 20 is a mobile phone.

  The information processing apparatus 10 interactively exchanges the question text and the answer to the question text with the user, narrows down the search target data based on the answer acquired from the user (the answer is a part of the search condition), and the result To the user. The provision of the question sentence and the acquisition of the answer are repeated a plurality of times depending on the result and the number of prepared questions.

  For example, FIG. 2 shows an example of a screen that the information processing apparatus 10 provides to the user apparatus 20. FIG. 2 is an explanatory diagram illustrating an example of a screen display of the user device. For example, in a restaurant search, it is assumed that a question sentence “Where is your desired place?” Is first provided, and the user transmits an answer “Tamachi” to that question. The information processing apparatus 10 extracts “restaurant in Tamachi” from the search target data and provides it to the user. Next, the information processing apparatus 10 may ask the cooking category “What do you want to eat?” Or “Would you like to have a parking lot?” You may ask questions.

  As described above, in the search system that interactively repeats the question and the answer and narrows down the target data, the order of the questions greatly affects the accuracy of narrowing down the search target data. Therefore, efficient control of the order of questions is a very important point in improving the quality of search services.

  Here, as one method for determining the order of questions, there is a method using an expected value. For example, Patent Document 1 discloses a system for determining an order of questions by calculating an expected value from the number of search results belonging to each class, with an answer to each question as a class in an interactive information search system applied to a television reservation. Is described. The expected value H (Cj) used here is obtained by the following mathematical formula (1).

  Here, Ntotal is the total number of program candidates, Nclass (i) is the number of programs belonging to each class, and I is the number of classes.

  In the method described in Patent Document 1, selection characteristics for each user are not taken into consideration, and questions are provided to all users in the same order. Also, since the answer to the question does not take into account the case of “not specified”, for example, when applied to the restaurant search described in the present embodiment, when the user often answers “not specified” the condition There was a problem that efficient narrowing down was not possible.

  Therefore, the information processing apparatus 10 according to the present embodiment described below calculates an expected value based on the user's answer history, and controls the order in which questions are provided based on the expected value.

<Functional Configuration of Information Processing Apparatus 10>
Next, a functional configuration of the information processing apparatus 10 for realizing the above functions will be described with reference to FIG. FIG. 4 is a block diagram illustrating a functional configuration of the information processing apparatus.

  The information processing apparatus 10 mainly includes an input / output control unit 11, a memory unit 12, a dialogue scenario storage unit 13, a dialogue history data storage unit 14, a search target data storage unit 15, and a dialogue control unit 16.

(Input / output control unit 11)
The input / output control unit 11 has a function of controlling input and output to the information processing apparatus 10. The input / output control unit 11 is connected to the dialogue control unit 16 and the communication network 40 (not shown here). In the present embodiment, the input / output control unit 11 provides the user with a screen including the question sentence selected by the dialogue control unit 16, and when the user inputs, for example, response data using the user device 20, the response Data is acquired and input to the dialogue control unit 16.

(Memory unit 12)
The memory unit 12 temporarily stores programs used by the input / output control unit 11 and the dialogue control unit 16, various calculation parameters, and the like. For example, the memory unit 12 stores an execution dialogue scenario 122, an order control flag 124, and an order control non-execution dialogue scenario list 126. The memory unit 12 may be a storage unit such as a ROM (Read Only Memory) and a RAM (Random Access Memory).

(Dialogue scenario storage unit 13)
The dialogue scenario storage unit 13 is a database that mainly stores dialogue scenario data 130, question sentence data 132, and option data 134. The dialogue scenario storage unit 13 has a function of storing and managing various data. When a request (for example, an SQL sentence) in which an extraction condition is set by the dialog control unit 16 is input, the dialog scenario storage unit 13 extracts data that matches the condition from the data in the database according to the set condition. It has a function of returning to the dialog control unit 16.

  Here, data stored in the dialogue scenario storage unit 13 will be described with reference to FIGS. 5, 6, and 7. FIG. 5 is an explanatory diagram showing an example of dialogue scenario data. FIG. 6 is an explanatory diagram showing an example of question sentence data. FIG. 7 is an explanatory diagram showing an example of option data.

  First, referring to FIG. 5, the dialogue scenario data 130 is data describing a dialogue scenario for controlling the flow of the dialogue. The dialogue ID 1301, the type 1302, the question sentence ID 1303, the option ID 1304, the database attribute 1305, and the next dialogue ID list. 1306 is mainly composed of a list of data. The dialogue ID 1301 is an identifier that uniquely identifies the dialogue. The type 1302 describes the type of dialogue. For example, the type 1302 describes a type such as “question” or “control”. In the case of control, the type of control is further described. The dialogue control unit 16 determines the next process according to this type. The question sentence ID 1303 is an identifier for uniquely identifying the question sentence data. The option ID 1304 is an identifier that uniquely specifies option data. A database attribute 1305 is a corresponding attribute in the database. The next dialog list 1306 is a list of dialog IDs that specify the next dialog. Here, “NULL” indicates that no data exists.

  The dialogue scenario data 130 is data described in advance by the designer in order to control the flow of dialogue. As described above, by performing the flow of the dialogue using the data separated from the program, even when a modification such as addition and deletion of a question sentence is added, the dialogue scenario data 130 is not modified. There is an effect that correction can be easily made by adding or deleting data.

  An example of a method for creating the dialogue scenario data 130 is as follows. For example, when asking questions in a fixed order, only one dialogue ID for specifying the next dialogue is described in the next dialogue ID list of one data. On the other hand, when a question is executed by selecting an optimal order for narrowing down, the type 1302 is set to “control: start sequence control”, and a list of conversation IDs in which the question to be narrowed down is described in the next conversation ID list. And the data describing the conversation ID for specifying the next conversation in the next dialog ID list 1306 are created in the next dialog ID list 1306, and the dialog for narrowing down is the next dialog. An ID list 1306 describes a dialogue ID of data for control. In order to specify the first dialog, the type 1302 is set to “control: dialog start”, and data describing the dialog ID for specifying the first dialog is prepared in the next dialog ID list 1306. In order to end the dialogue, data having the type 1306 as “control: dialogue end” is prepared, and the last dialogue describes the dialogue ID of the data whose type is “control: dialogue end” in the next dialogue ID list 1306. To do. By creating such data, the designer can create a scenario for controlling the interaction.

  Next, referring to FIG. 6, the question sentence data 132 mainly includes a question sentence ID 1321 and a question sentence 1322. The question sentence ID 1321 is an identifier that uniquely identifies the question. The question sentence 1322 is a text indicating the content of the question.

  Next, referring to FIG. 7, the option data 134 mainly includes an option ID 1341 and an option list 1342. The option ID 1341 is an identifier that uniquely specifies an option. The option list 1342 is a list that lists options.

(Dialogue history data storage unit 14)
The dialogue history data storage unit 14 is a database that stores dialogue history data 140. When a request (for example, an SQL statement) for which a condition is set by the dialog control unit 16 is input, the dialog history data storage unit 14 performs processing according to the set condition. For example, in the case where an extraction command is described in the request, it has a function of extracting data that matches the condition and returning it to the dialog control unit 16. Further, for example, when an extraction command for the number of cases is described in the request, it has a function of counting the number of pieces of data that meet the conditions and returning the data to the dialogue control unit 16.

  Here, an example of data stored in the dialogue history data storage unit 14 will be described with reference to FIG. FIG. 8 is an explanatory diagram showing an example of the conversation history data. The conversation history data 140 mainly includes a date 1401, a user ID 1402, a question sentence ID 1403, and an answer 1404. The conversation history data 140 is information that accumulates what kind of selection action is performed when a specific user uses a service provided by the information processing apparatus 10. One piece of data is generated when the user selects an answer to the question provided by the information processing apparatus 10. The date and time 1401 is the date and time when the data was generated. The user ID 1402 is an identifier that uniquely identifies the user. The question sentence ID 1403 is an identifier that uniquely identifies the question provided by the information processing apparatus 10. The answer 1404 is an answer input by the user with respect to the question specified by the question sentence ID 1403. Here, the dialogue history data 140 includes “not specified” as the reply 1404.

  The information processing apparatus 10 according to the present embodiment uses the conversation history data 140 to control the order of questions suitable for each user, and to narrow down search target data efficiently. A mode of using the dialogue history data 140 will be described later.

(Search target data storage unit 15)
The search target data storage unit 15 is a database that stores the search target data 150. When a request (for example, an SQL statement) for which a condition is set by the dialog control unit 16 is input, the search target data storage unit 15 performs processing according to the set condition. For example, in the case where an extraction command is described in the request, it has a function of extracting data that matches the condition and returning it to the dialog control unit 16. Further, for example, when an extraction command for the number of cases is described in the request, it has a function of counting the number of pieces of data that meet the conditions and returning the data to the dialogue control unit 16.

  Here, an example of the search target data 150 stored in the search target data storage unit 15 will be described with reference to FIG. FIG. 9 is an explanatory diagram of an example of search target data. The search target data 150 is a set of information related to a specific target object. In the present embodiment, restaurant information will be described as an example. The search target data 150 is composed of a list of one data consisting of a plurality of sets of combinations of data IDs 1501 and attributes and attribute values. Here, five types of combinations of attributes and attribute values are included in one piece of data, but the present invention is not limited to this example. Here, the attributes include, for example, “NAME” indicating the restaurant name, “CAT” indicating the restaurant category, “LOC” indicating the restaurant location, and “whether or not the restaurant has a parking lot”. “PARK” and “PET” indicating whether or not the restaurant is allowed to enter the pet. The attribute value is a specific value of the corresponding attribute.

(Dialogue control unit 16)
The dialogue control unit 16 is connected to the input / output control unit 11, the memory unit 12, the dialogue scenario storage unit 13, the dialogue history data storage unit 14, and the search target data storage unit 15, and controls the operation of each unit of the information processing apparatus 10. To interact. The dialogue control unit 16 executes, for example, processing performed in each block of FIGS. Detailed contents of the control of the dialogue control unit 16 will be described later. At this time, the dialogue control unit 16 acquires or generates information required during execution of each process such as the execution dialogue scenario 122, the sequence control flag 124, the sequence control non-executed dialogue scenario list 126, and the like. 12 is temporarily stored and used. Further, the dialogue control unit 16 inputs information to be provided to the user to the input / output control unit 11 and receives the information acquired by the input / output control unit 11 from the user. Furthermore, the dialogue control unit 16 inputs a request including conditions for acquiring various information to the dialogue scenario storage unit 13, the dialogue history data storage unit 14, and the search target data storage unit 15, and obtains desired information. To get.

  The function blocks of the information processing apparatus 10 have been described above. Here, the input / output control unit 11 and the dialogue control unit 16 may be realized by an arithmetic processing device such as a CPU, for example. The input / output control unit 11 and the dialogue control unit 16 control the operation in the apparatus according to various programs. In FIG. 4, the dialogue scenario storage unit 13, the dialogue history data storage unit 14, and the search target data storage unit 15 are all included in the information processing apparatus 10, but are not limited thereto. A form using a database of an external device may be used. In this case, the dialogue control unit 16 inputs a request including an information acquisition condition to each database via the input / output control unit 11 and the communication network 40.

<Dialogue control action>
Next, the overall flow of the dialogue control operation of the information processing apparatus 10 according to the present embodiment will be described with reference to FIG. FIG. 10 is a flowchart showing the overall flow of dialog control. Details of each process shown in FIG. 10 will be described later with reference to FIGS.

  First, the dialogue control unit 16 executes dialogue initial processing (S102). The detailed operation of the dialog initial process, which will be described later, includes an operation from when the user accesses the information processing apparatus 10 to receive service provision until the dialog control unit 16 acquires a dialog scenario.

  Then, the dialogue control unit 16 executes each process according to the acquired scenario type. Therefore, the scenario type is determined next (S104). In the determination of step S104, when the scenario type is “question”, a dialogue process (S106) including a process of providing a question to the user and acquiring the answer, and an object corresponding to the obtained answer Data search processing (S108) is executed.

  In the determination in step S104, if the scenario type is “start sequence control”, the user is requested to select one of the next scenario data lists described in the “next conversation ID list” of the scenario data. An order control process (S110) for controlling the order of questions to be provided is executed. This order control process includes a process for determining the order of questions and a dialog process with the user according to the order.

  In addition, after the dialogue process in step S106, the target data search process in step S108, and the order control process in step S110 are executed, and in the determination in step S104, the scenario type is “end of order control”. Next, the dialogue control unit 16 sets the dialogue ID described in the “next dialogue ID list” in the scenario data to the dialogue ID for acquiring the dialogue ID (S112). Then, the dialogue control unit 16 executes a dialogue scenario data acquisition process based on the set dialogue ID (S114). Thereafter, the process returns to the process of determining the scenario type in step S104 based on the acquired scenario data.

  If the scenario type is “dialog end” in the determination in step S104, the dialog process ends.

  Next, details of dialogue control of the information processing apparatus 10 according to the present embodiment will be described with reference to FIGS. FIGS. 11 to 19 are sub-flowcharts showing a part of the dialogue control shown in FIG. Specifically, FIG. 11 is a flowchart showing the flow of the dialog initial process. FIG. 12 is a flowchart showing the flow of the dialogue scenario data acquisition process. FIG. 13 is a flowchart showing the flow of dialogue processing. FIG. 14 is a flowchart showing the flow of the question sentence data acquisition process. FIG. 15 is a flowchart showing the flow of option data acquisition processing. FIG. 16 is a flowchart showing the flow of target data search processing. FIG. 17 is a flowchart showing the sequence control process. FIG. 18 is a flowchart showing the flow of the conversation history data count process. FIG. 19 is a flowchart showing the flow of target data count processing.

  The detailed flow described below will be described in the case where it is executed along the scenario data 130 shown in FIG.

(Initial dialogue processing)
Referring to FIG. 10, first, an initial dialog process is executed (S102). The detailed operation of the dialog initial process is shown in FIG. Referring to FIG. 11, the dialog initial process starts when the user device 20 first accesses the information processing device 10 (S202). Then, the input / output control unit 11 acquires the user ID and instructs the dialogue control unit 16 to start dialogue processing (S204). Here, the method by which the input / output control unit 11 obtains the user ID includes, for example, a method for allowing the user to input the user ID on the screen, a method for obtaining the individual identification number of the mobile phone, a method using Cookie, and the like. And may be obtained by any method. For example, a user ID “U0123” is acquired here.

  Next, the dialogue control unit 16 sets a specific ID indicating start data as a dialogue ID to be transferred to the dialogue scenario data acquisition process (S206). Here, the specific dialogue ID is, for example, “SD000” in FIG. Then, the dialogue control unit 16 executes dialogue scenario data acquisition processing (S208).

(Interaction scenario data acquisition process)
Details of the dialogue scenario data acquisition process are shown in FIG. Referring to FIG. 12, the dialogue control unit 16 transmits a scenario data acquisition condition including the dialogue ID set in the immediately preceding step S206 to the dialogue scenario storage unit 13 (S302). The dialogue scenario storage unit 13 receives the scenario data acquisition condition, acquires the scenario data that matches the scenario data acquisition condition, and transmits it to the dialogue control unit 16 (S304). For example, the dialogue control unit 16 has the dialogue ID 1301 of “SD000”, the type 1302 of “control: dialogue start”, the question sentence ID 1303 of “NULL”, the option ID 1304 of “NULL”, the database attribute 1305 of “NULL”, and the next dialogue. One dialogue scenario data whose ID list 1306 is “SD001” is acquired (S306).

  Then, the dialogue control unit 16 sets “SD001”, which is data in the next dialogue ID list 1306 column of the scenario data with the obtained dialogue ID “SD000” as the ID to be transferred to the dialogue scenario data acquisition process ( S210). Then, the dialogue control unit 16 executes dialogue scenario data acquisition processing based on the dialogue ID “SD001” (S212). The details of the dialogue scenario data acquisition process are the same as steps S302 to S306 in FIG. Here, the dialogue control unit 16 has the dialogue ID 1301 “SD001”, the type 1302 “question”, the question sentence ID 1303 “Q001”, the option ID 1304 “C001”, the database attribute 1305 “LOC”, and the next dialogue ID list. One dialog scenario data in which 1306 is “S010” is acquired. This completes the dialog initial processing.

(Interactive processing)
Referring again to FIG. 10, the next process is determined according to the scenario data type 1302 acquired in the dialog initial process step S102. Therefore, the scenario type 1302 is determined (S104). Here, since the type of the scenario acquired last in step S102 is “question”, the dialogue control unit 16 executes dialogue processing (S106).

  Details of the dialogue processing will be described with reference to FIG. Referring to FIG. 13, first, the dialogue control unit 16 sets the dialogue ID “SD001” in the execution dialogue scenario 122 of the memory unit 12 (S402). Then, the dialogue control unit 16 executes a question sentence data acquisition process (S404). The details of the question sentence data acquisition process will be described with reference to FIG.

(Question sentence data acquisition process)
First, the dialogue control unit 16 transmits the question sentence data acquisition condition including the question sentence ID “Q001” included in the currently executed dialogue scenario to the dialogue scenario storage unit 13 (S312). And the dialogue scenario memory | storage part 13 will acquire question sentence data based on conditions, if a question sentence data acquisition condition is received, and will reply to the dialogue control part 16 (S314). In the question sentence data acquired here, the question sentence ID is “Q001” and the question sentence is “Where is the desired place?”. Then, the dialogue control unit 16 acquires the question sentence data (S316).

(Option data acquisition process)
Referring to FIG. 13 again, the dialogue control unit 16 next executes option data acquisition processing (S406). Refer to FIG. 15 for details of the option data acquisition process. First, the dialogue control unit 16 transmits an option data acquisition condition including an option ID included in the currently executed dialogue scenario to the dialogue scenario storage unit 13 (S322). Then, the dialogue scenario storage unit 13 acquires option data based on the received condition and sends it back to the dialogue control unit 16 (S324). The option data acquired here has the option ID “C001” and the option list “Ginza, Shinjuku, Shibuya, Ikebukuro, Tamachi, Yokohama”. Then, the dialogue control unit 16 acquires the option data (S326).

  Then, the dialogue control unit 16 sets the acquired question sentence data and option data as output data to be passed to the input / output control unit 11 (S408). At this time, it is determined whether or not there is search target data as a search result to be displayed (S410). Here, when there is search target data that is a search result to be displayed, that is, when the interactive process is executed for the second time or later, and there is a search result of the previous interactive process, the data that is the search result Is set to output data to be passed to the input / output control unit 11 (S412). Then, the dialogue control unit 16 transmits a command for executing the dialogue start to the input / output control unit 11 together with the output data (S414).

In response to the instruction from the dialogue control unit 16 in step S414, the input / output control unit 11 generates an output screen based on the output data. At this time, the input / output control unit 11 adds data such as “no designation” as an option (S416). Then, the user terminal 20 receives the screen generated by the input / output control unit 11 and displays it on the display unit of the user terminal 20. At this time, a screen as shown in FIG. 2 is displayed on the display unit of the user terminal 20, for example. For example, the question “Where is your preferred location?”
And, the option “Ginza, Shinjuku, Shibuya, Ikebukuro, Tamachi, Yokohama” is displayed with a pull-down menu or numbered, and the input is waited for. Alternatively, a screen as shown in FIG. 3 may be displayed. Answers can be freely entered in text and matched with options in internal processing. At this time, voice input / output may be used.

  In this case, since it is the first interactive process, a question sentence and an option for it are displayed. If it is the second and subsequent dialogue processing, the search result in the previous dialogue processing may be displayed. On this screen, the user performs an operation of inputting an answer to the question sentence. Then, the user terminal 20 acquires the user operation and transmits it to the input / output control unit 11 (S418). For example, it is assumed here that the answer is “Tamachi”.

  When obtaining the answer data from the user terminal 20, the input / output control unit 11 transmits the obtained answer data to the dialogue control unit 16 (S420). When the dialogue control unit 16 obtains the answer data, the dialogue history data sets the question sentence ID corresponding to the answer data (here, the question sentence ID in the conversation scenario being executed) and the user ID. The data is transmitted to the storage unit 14 (S424). The dialogue history data storage unit 14 updates the dialogue history data based on the received data (S426). This completes the dialogue process.

(Target data search process)
Here, returning to FIG. 10 again, next, the target data search process is executed (S108). FIG. 16 is referred to for the detailed flow of the target data search process. First, the dialogue control unit 16 transmits a target data search condition including an attribute and an attribute value to the search target data storage unit 15 (S332). Here, the attribute is an attribute described in the database attribute column of the dialog scenario being executed. Here, referring to the database attribute 1305 of the dialogue ID “SD001”, it can be seen that it is “LOC”. The attribute value is a user's answer in the immediately preceding dialog process. Here, the answer “Tamachi” acquired in step S418.

  Then, the search target data storage unit 15 searches the search target data based on the received attribute and attribute value, extracts the corresponding target data, and transmits it as a search result to the dialog control unit 16 (S334). That is, the search target data storage unit 15 extracts data having the attribute “LOC” and the attribute value “Tamachi”, and transmits the data to the dialogue control unit 16 as a search result. The dialogue control unit 16 obtains the search result by receiving the search result (S336).

Returning to FIG. 10 again, the search target data as restaurant information is narrowed down by the condition that the place is “Tamachi” by the target data search process of step S108 described above. Thereby, the process according to the scenario data with the dialogue ID “SD001” is completed. In order to obtain the next scenario data, the dialogue control unit 16 obtains the next dialogue ID list “SD010” from the column of the next dialogue ID list 1306 of the scenario data having the dialogue ID “SD001”. (S112). Then, scenario data whose dialogue ID is “SD010” is acquired (S114). Here, the dialogue control unit 16 has a dialogue ID 1301 of “SD010”, a type 1302 of “control: start sequence control”, and a question sentence ID 1303 of “NULL”.
One scenario data with the option ID 1304 “NULL”, the database attribute 1305 “NULL”, and the next dialogue ID list 1306 “SD011, SD012, SD013” is acquired.

  Here again, the dialogue control unit 16 determines the type of scenario data (S104). Since the type of scenario data acquired at this time is “control: start sequence control”, the dialog control unit 16 next executes sequence control processing (S110).

(Sequence control processing)
Refer to FIG. 17 for details of the order control process. First, the dialogue control unit 16 turns on the order control flag 124, and sets “SD011, SD012, SD013”, which is the next dialogue ID list in the scenario data obtained at the present time, to the obtained dialogue ID (S502). Then, scenario data whose dialogue ID is “SD011, SD012, SD013” is acquired (S504). Here, the details of the scenario data acquisition processing have already been described with reference to FIG.

  Next, the dialogue control unit 16 sets the dialogue IDs of all the dialogue scenario data acquired in step S504 in the sequence control unexecuted scenario list 126. The sequence control non-executed scenario list 126 is set with scenario data that needs to be sequence controlled and has not been provided to the user. That is, the dialogue control unit 16 first registers all target scenario data, and when the sequence control is completed, deletes it from the sequence control unexecuted scenario list 126 to identify the sequence control target data. I can do it. Therefore, at this point, the dialogue control unit 16 sets “SD011, SD012, SD013” in the sequence control unexecuted scenario list 126.

  Then, the dialogue control unit 16 executes option data acquisition processing based on “C011, C012, C013” described in each option ID 1304 of the acquired scenario data “SD011, SD012, SD013” (S508). That is, in such an example, the option ID 1341 is “C011”, the option list 1342 is one option data “Japanese, Chinese, Western, Cafe”, the option ID 1341 is “C012”, and the option list is “C012”. One option data “Yes, No”, and one option data whose option ID 1341 is “C013” and whose option list 1342 is “Yes, No” are acquired.

(Dialogue history data count processing)
Next, the dialogue control unit 16 executes a dialogue history data count process (S510). FIG. 18 is referred to for details of the dialog history data count process. The information processing apparatus 10 according to the present embodiment considers not only the number of search results but also a selection history for each user as a condition for controlling the order of questions. For this reason, the dialogue control unit 16 executes a process of counting dialogue history data that is basic data for calculating expected values.

  The dialog control unit 16 first transmits a dialog history data count condition including an option, a user ID, and a question sentence ID to the dialog history data storage unit 14. At this time, “unspecified” is added to the options. The dialogue history data storage unit 14 counts the number of dialogue history data based on the received dialogue history data count condition (S604), and transmits the count data to the dialogue control unit 16 (S606). The dialogue control unit 16 acquires count data from the dialogue history data storage unit 14 (S608).

  For example, the dialogue control unit 16 sends the options “Japanese food”, “Chinese food”, “Western food”, “cafe”, “unspecified”, user ID “U0123”, and question sentence ID “Q011”. “0”, “Chinese” 5 items, “Western food” 2 items, “Cafe” 3 items, “Unspecified” 0 items are acquired. Such data indicates how many times the corresponding user has selected each option in the past.

  Similarly, the dialogue control unit 16 sends the options “Yes”, “No”, “No designation”, the user ID “U0123”, and the question sentence ID “Q012”, respectively, Get “none” 0 items and “unspecified” 1 item. Similarly, the dialogue control unit 16 sends the choices “possible”, “impossible”, “not specified”, the user ID “U0123”, and the question sentence ID “Q013”. Acquire 0 “impossible” and 5 “unspecified”.

  The dialogue control unit 16 determines whether or not the received count data is a sufficient number to be used as basic data for selecting a question sentence (S610). The dialogue control unit 16 may perform such determination by providing a predetermined threshold value (hereinafter referred to as a history number lower limit threshold value) for the number of count data. For example, if the history number lower limit threshold is 10, the count data for the question sentence ID “Q013” is 5 in the above count data. At this time, since the number of cases is less than the history number lower limit threshold, the complementing process shown in steps S612 to S620 of FIG. 18 is executed.

  The dialogue control unit 16 transmits the dialogue history data count condition including the option and the question ID to the dialogue history data storage unit 14. Here, unlike step S602, the user ID is not included in the dialog history data count condition. Then, the dialogue history data storage unit 14 counts the dialogue history data without limiting the user ID (S614). That is, the dialogue control unit 16 transmits a dialogue history data count condition including the choices “possible”, “impossible”, and the question sentence ID “Q013”, and the dialogue history data storage unit 14 includes five “possible” and “impossible”. “0” and “not specified” 495 count data are acquired.

  The dialogue history data storage unit 14 transmits the acquired count data to the dialogue control unit 16 (S616). When the dialog control unit 16 receives the count data from the dialog history data storage unit 14 (S618), the count data acquired based on the user ID using the count data is counted based on the dialog history including other IDs. It complements with data (S620). In other words, in order to supplement less than 10 cases with the count data, the count data of “Yes”, “No”, “Not specified”, and 495 counts is converted so that the total count becomes 5 "Acceptable" 0.05, "Not allowed" 0, "Not specified" 4.95 added, "Yes" 0.05, "Not allowed" 0, "Not specified" 9.95 And

  Although not shown, when the number of acquired conversation history data is sufficient, the number of cases may be counted while being further limited by adding conditions. For example, when the count data of the acquired dialogue history data is greater than or equal to a predetermined threshold (hereinafter referred to as history history number upper limit threshold), the dialogue control unit 16 sets at least one of time, place, date, and day of the week. A question sentence to be provided to the user may be selected using the number of conversation history data further narrowed down. Specifically, the dialogue control unit 16 restricts the dialogue history data to the data of the same time zone, the data of the same day of the week, the data of the latest date and time, or the location of the mobile phone can be acquired. In some cases, the location may be limited using this information. The dialog history data count processing step S510 is now complete.

(Target data count processing)
Referring to FIG. 17 again, next, the dialogue control unit 16 executes a target data count process (S512). For details of the target data count process, refer to FIG. First, the dialogue control unit 16 transmits a target data count condition including a list of attributes and attribute values to the search target data storage unit 15 (S702). The search target data storage unit 15 counts the target data based on the received target data count condition (S704).

  For example, the dialogue control unit 16 sets the target data count condition in which the already specified attribute “LOC” is “Tamachi” and the attribute “CAT” is “Japanese food”, “Chinese food”, “Western food”, “Cafe”. The search target data storage unit 15 obtains count data of “Japanese food” 10 items, “Chinese food” 25 items, “Western food” 15 items, and “Cafe” 20 items. The dialogue control unit 16 transmits a target data count condition in which the attribute “LOC” is “Tamachi” and the attribute “PARK” is “present” or “none”, and the search target data storage unit 15 Count data of “Yes”, 10 cases, “No”, 90 cases, and “NULL” 0 cases are acquired. The dialogue control unit 16 transmits a target data count condition in which the attribute “LOC” is “Tamachi” and the attribute “PET” is “possible” or “impossible”. The search target data storage unit 15 Count data that is “possible” 5 cases, “impossible” 90 cases, and “NULL” 5 cases is acquired.

(Calculation of expected value)
Next, the dialogue control unit 16 performs, for each dialogue scenario data described in the sequence control unexecuted dialogue scenario list 126, based on the obtained number of dialogue history data of each option and the number of target data. Then, the expected value of the number of narrowed down cases by dialogue is calculated (S514).

  In the present embodiment, the expected value E (SDi) of the number of narrowed-down cases of the dialogue scenario data SDi (i = 1, 2,...) Is calculated using the following formula (2).

  Here, each option of the dialogue scenario data SDi is Cij (j = 1, 2,...), The number of dialogue history data by the choice Cij is Nl (Cij), and the dialogue history of “unspecified” in the dialogue scenario data SDi. It is assumed that the number of data items is N10 and the number of target data items with the option Cij is Nd (Cij).

  For example, the expected value E of narrowing down by each scenario data “SD011”, “SD012”, “SD013” is calculated as follows.

  As described above, the order in which the question is provided to the user is determined based on the expected value obtained in the expected value calculation step of step S514. In the present embodiment, questions are provided in order of increasing expected values. Accordingly, the dialogue control unit 16 next selects the dialogue ID having the smallest expected value (S516). Here, since E (SD012) is the minimum, the dialogue control unit 16 selects SD012.

  Next, the dialogue control unit 16 executes dialogue processing based on the dialogue ID selected in step S516 (S518). The detailed operation of the dialogue processing has been described above with reference to FIG. The dialogue control unit 16 provides a question included in the scenario data specified by the dialogue ID “SD012” to the user through this dialogue processing, and obtains the answer. For example, it is assumed here that a question “Would it be better to have a parking lot?” Was provided and an answer “Yes” was obtained. Next, the dialogue control unit 16 executes target data search processing based on the acquired answer (S520). Then, the dialogue control unit 16 deletes SD012, which is the dialogue ID selected here, from the sequence control non-executed dialogue scenario list (S522).

  Here, the dialogue control unit 16 determines whether or not the number of search target data narrowed down as a result of the search process executed in step S520 is equal to or less than a predetermined threshold (hereinafter referred to as a result number threshold). S523). Here, the result number threshold may be determined based on, for example, the number of cases in which the user can easily view the results, or the number of cases in which the user can easily find desired information from the results. For example, the threshold value for the number of results may be determined in relation to a screen operation for the user to view the results, or it may be difficult for the user to perform a scroll operation on the screen of the mobile phone based on the number of cases that fit within the screen of the mobile phone. It may be determined within the range. As a result of the determination, when the result number is equal to or smaller than the threshold value, the dialogue control unit 16 ends the order control process.

  Here, when the result of determination in step S523 is that the result is not less than the result number threshold, the dialog control unit 16 determines whether or not the order control unexecuted dialog list after execution of deletion in step S522 is zero. (S524). At this time point, “SD011” and “SD013” are described in the sequence control non-executed dialogue scenario list. Next, the dialog control unit 16 has one sequence control non-executed dialogue scenario list. It is determined whether or not. Since two cases are described at this point, the process returns to step S512.

  Here, the dialogue control unit 16 again executes the target data count process for the remaining two scenario lists, and calculates the expected value of the number of narrowed down cases (S512 to S514). At this time, for example, in the target data counting process (S512), the already designated attribute “LOC” is “Tamachi”, “PARK” is “Yes”, and the attribute “CAT” is “Japanese food”, “ The target data count conditions of “Chinese”, “Western food”, and “cafe” are transmitted to obtain count data. In addition, the dialogue control unit 16 sets the target data in which the specified attribute “LOC” is “Tamachi”, “PARK” is “Yes”, and the attribute “PET” is “Yes” and “No”. The count condition is transmitted, and the count data is acquired for each. Based on the count data acquired here, the dialogue control unit 16 similarly executes the processing of step S514 to step S516. For example, when the dialogue ID having the minimum expected value is “SD011” at this time, the dialogue control unit 16 executes the processes of steps S518 to S522 based on the selected dialogue ID.

  Then, in step S523, the dialogue control unit 16 determines again whether or not the number of pieces of search target data narrowed down by the target data search process executed based on the dialogue ID “SD011” is equal to or less than the result number threshold. .

  If the number of search target data after narrowing down is not less than or equal to the result number threshold, the dialog control unit 16 determines whether or not the order control unexecuted dialog scenario list is zero in step S524. Here, since there is only one remaining, the dialogue control unit 16 determines whether or not there is one sequence control unexecuted dialogue scenario list (S526). Since one item is described at this time, the dialogue control unit 16 executes the processing of Step S518 to Step S522 based on the remaining dialogue ID. As a result of the above processing, the order control unexecuted scenario list becomes 0, so the order control processing is ended by the determination in step S524.

  Here, “SD019” is described in the next dialogue ID list of the scenario data with the dialogue ID “SD013”. Therefore, the dialogue control unit 16 sets “SD019”, which is a list of next dialogue IDs, to the obtained dialogue ID (S112). Then, the dialogue control unit 16 executes a scenario data acquisition process based on the set dialogue ID (S114). That is, the dialogue control unit 16 has the dialogue ID 1301 of “SD019”, the type 1302 “control: sequence control end”, the question sentence ID 1303 “NULL”, the option ID 1304 “NULL”, the database attribute 1305 “NULL”, and the next dialogue ID. One scenario data of list 1306 “SD999” is acquired.

  Then, the dialogue control unit 16 executes the scenario type determination process in step S104 based on the acquired scenario data. At this time, since the scenario type is “sequential control end”, the dialogue control unit 16 next sets “SD999”, which is a list of next dialogue IDs, to the obtained dialogue ID (S112). Then, the dialogue control unit 16 executes dialogue scenario data acquisition processing (S114) based on this dialogue ID. That is, the dialogue control unit 16 has the dialogue ID 1301 of “SD999”, the type 1302 “control: dialogue termination”, the question sentence ID 1303 “NULL”, the option ID 1304 “NULL”, the database attribute 1305 “NULL”, and the next dialogue ID list. One scenario data 1306 “NULL” is acquired.

  Then, the dialogue control unit 16 executes the scenario type determination process in step S104 based on the acquired scenario data. At this point in time, since the scenario type is “dialog end”, the dialog control unit 16 ends the dialog control.

<Example of effects>
Heretofore, the functional configuration related to dialog control and the dialog control operation of the information processing apparatus 10 according to the present embodiment have been described. According to such a configuration, when controlling the dialog order, the dialog control unit 16 calculates the expected value of the number of narrowed down cases based on the number of user's dialog history data and the number of search target data, and the expected value is Execute conversations based on minimal interaction scenarios. Thereby, since the user's selection history can be reflected in the selection of the interaction order, there is an effect that the user's preferences and restrictions can be reflected in the determination of the interaction order.

  Also, when there is sufficient dialogue history data, the dialogue control unit 16 limits the user's dialogue history data using conditions such as the same time zone, the same day of the week, and the most recent date and time. Alternatively, the location is limited when the location of the user can be acquired using the function of the mobile phone or the like. Thereby, there is an effect that it is possible to execute optimum narrowing control in accordance with an action pattern according to the user's time and day of the week.

  When the dialogue history data is insufficient, the dialogue control unit 16 supplements the deficient data using the history data of other people. Here, it is possible to create reliable data while retaining the characteristics of each user as much as possible by complementing only the shortage with other person's history data instead of using the data of others. .

  Further, the dialogue control including the provision of the question sentence and the acquisition of the answer is continued until the number of the search target data after narrowing down becomes a predetermined result number threshold value or until there are no candidate question sentences. Thereby, it is possible to increase the possibility of controlling the dialog so that the number of search results is equal to or less than the desired number.

  Further, the dialogue history data includes “no designation” as an answer. As a result, it is possible to calculate the expected value considering the case where an answer that has not been considered in the past is not specified, and to determine the order in which the question sentences are provided based on more realistic and accurate data. It becomes like this.

  The preferred embodiments of the present invention have been described in detail above with reference to the accompanying drawings, but the present invention is not limited to such examples. It is obvious that a person having ordinary knowledge in the technical field to which the present invention pertains can come up with various changes or modifications within the scope of the technical idea described in the claims. Of course, it is understood that these also belong to the technical scope of the present invention.

  For example, in FIG. 9, the restaurant location information stores words indicating place names, such as “Tamachi” and “Shibuya”, but the present invention is not limited to this example. For example, an address may be stored. In this case, a matching process with an address is separately required based on an answer to a question about a place.

  In this specification, the steps described in the flowcharts are executed in parallel or individually even if they are not necessarily processed in time series, as well as processes performed in time series in the described order. Including processing to be performed. Further, it goes without saying that the order can be appropriately changed even in the steps processed in time series.

DESCRIPTION OF SYMBOLS 10 Information processing apparatus 11 Input / output control part 12 Memory part 13 Dialogue scenario storage part 14 Dialogue history data storage part 15 Search object data storage part 16 Dialogue control part

Claims (7)

A dialogue scenario storage unit for storing a question sentence to be provided to a user;
A dialogue history data storage unit for storing dialogue history data which is an answer history for each user with respect to the question sentence;
Based on the number of the conversation history data and the number of search target data, select a question sentence to be provided to the user, and a dialog control unit that narrows down the search target data based on the user's answer to the question sentence;
Input / output control unit for controlling execution of provision of the question sentence to the user, acquisition of an answer to the question sentence from the user, and provision of the search target data narrowed based on the answer to the user When,
An information processing apparatus comprising:   The dialog control unit is a search that can be narrowed down by the question text based on the number of search target data corresponding to each option of the answer to the question text and the number of the dialog history data corresponding to each option. The information processing apparatus according to claim 1, wherein an expected value of the number of target data is calculated, and a question sentence to be provided to the user is selected from a plurality of the question sentences based on the expected value. .   The said dialog control part repeats selection of a question sentence until the number of the said search object data provided to the said user becomes below a predetermined result number threshold value, The Claim 1 or 2 characterized by the above-mentioned. Information processing device.   The information processing apparatus according to claim 1, wherein the dialogue history data includes data when an answer to the question sentence is not designated.   When the number of conversation history data corresponding to the user is equal to or less than a predetermined history number lower limit threshold, the conversation control unit provides the user with the number of conversation history data not limited to the user. The information processing apparatus according to claim 1, wherein a question sentence is selected.   The dialogue control unit further narrows down using the time, place, date, and day of the week when the number of dialogue history data corresponding to the user is equal to or greater than a predetermined history number upper limit threshold. The information processing apparatus according to claim 1, wherein a question sentence to be provided to the user is selected using the number of conversation history data.   A dialogue scenario storage unit that stores question sentence data provided to a user, a dialogue history data storage unit that stores dialogue history data that is an answer history for each user with respect to the question sentence, and a question sentence provided to the user An interactive control unit that selects and narrows search target data based on the user's answer to the question sentence, provides the question sentence to the user, obtains an answer to the question sentence from the user, and The dialogue control unit of the information processing apparatus having an input / output unit that executes provision of the search target data to the user that is narrowed down based on the number of the dialog history data and the number of the search target data And selecting a question sentence to be provided to the user.

Images