JP4096570B2 - Information system, terminal, information acquisition method, program - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
It relates to information systems.
[0002]
[Prior art]
As an information system for searching and notifying information, a center and a terminal are provided. Information is classified and stored in the center for each category, and a list of categories is provided to the user of the terminal in response to a request from the terminal. There is known an information system that provides information searched for to a user by outputting to the terminal and notifying information that matches the category selected by the user from the list. In addition, there is known an information system in which a phrase included in information stored in a center is notified from a terminal by performing a full text search or the like and provided to a user.
[0003]
[Problems to be solved by the invention]
In such an information system, a category is selected from a user's terminal or a simple keyword is input, and information is searched based on the selected category / keyword. For example, search on the Internet In Web search using an engine or the like, there is a problem that many pages not intended by the user are hit. Therefore, it takes a lot of time and labor to discover the target information for the user, which is a burden on the user.
[0004]
Thus, for example, a method of searching for information corresponding to an ambiguous question in the natural language from the user by understanding the intention or meaning of the question from the natural language input by the user can be considered. However, in order to respond to such vague questions of users, it is necessary to perform semantic understanding processing at the center or terminal. Therefore, for example, there is a problem that it is difficult to perform such processing using a device having limited resources such as a car navigation device or a portable terminal as a terminal.
[0005]
SUMMARY OF THE INVENTION An object of the present invention is to provide an information system or the like that can eliminate the processing for understanding the meaning and can easily obtain the answer to the question in the natural language acquired from the questioner. And
[0006]
[Means for Solving the Problems and Effects of the Invention]
  According to the information system according to claim 1, which has been made to solve the above-described problems, a plurality of categories for classifying questions and the strength of each respondent in each category are stored in association with each other. And respondent information including at least the respondent's location information indicating the area where the respondent is good at, and the natural language questions from the questioner, the category of the question, and the position related to the question Information, and at least additional information from the questioner for the question including the location information,Respondent candidates are determined based on the respondent information and additional information, the respondent candidate's goodness is extracted from the goodness of each respondent, and the respondent is based on the goodness of the respondent candidate in the question category Rank the candidates, and at least the top oneRespondent to whom the question is outputWhenDecideMemorize the respondent information about the determined respondentThe question is output to the respondent who is the output target. Then, a natural language answer from the respondent to the question is acquired, and the answer is output to the questioner. In particular, when the questioner does not explicitly specify position information related to the question, information indicating the current position of the questioner is acquired as position information related to the question. Therefore, the questioner can obtain a natural language answer to the question in his natural language. That is, it is possible to provide an information system that can respond to an ambiguous inquiry from a user. Moreover, by doing in this way, it is not necessary to perform processing for understanding the meaning in the information system. Therefore, the answer to the question in the questioner's natural language can be easily notified to the questioner in the natural language.
[0007]
Moreover, the respondent to whom the question acquired from the questioner is output is determined based on the level of expertise in each category for each respondent. Therefore, for example, a question can be output only to a respondent who is good at the question category from among a large number of respondents, so that there is a high possibility that an accurate answer to the question can be quickly obtained. In addition, it is not necessary to send a question to all of many respondents, and the processing load on the information system can be reduced. Each question and answer is data in a predetermined data format.
[0009]
  AlsoIn addition to categories, additional information is acquired, and respondents are determined by considering the additional information.RuyoSea urchinBecauseFurthermore, there is a high possibility that an accurate answer to the question can be quickly obtained. In addition, since it is possible to further narrow down the respondents, the processing load can be further reduced. Therefore, it becomes possible for the questioner to easily obtain answers to questions in natural language acquired from the questioner..
[0010]
  In this way, as a method of determining respondents with additional informationWithTo make a decision based on additional information and stored information about respondentsLikeFor,For example, it is possible to determine only the respondent who includes the matching information by comparing the stored information regarding the respondent and the additional information as the output respondent.
[0011]
  And as such additional informationIsUse location informationYes,As information about respondentsTimeThe position information indicating the area where the respondent is good can be used. For example, the area where the respondent knows well. In this way, when the content of the question is information related to a specific area, the respondent can be accurately determined, and the questioner can easily obtain an answer to the question. The region where the respondent is good may be determined from information such as the address of the respondent, for example, or information regarding the region where the respondent is good may be acquired in advance and the acquired information may be used. . Also,If the questioner does not explicitly specify the location information related to the question, the information indicating the current location of the questioner is acquired as location information related to the question.Accurate information near the current position can be easily obtained..
In addition, as described in claim 2, if information indicating a real-time current position is acquired using a GPS provided in a questioner's terminal or the like, information in the vicinity of the current position can be easily obtained. .
[0012]
  Further, as additional information, the claim3As shown in FIG. 4, the respondent designation information is included, and the respondent designated by the respondent designation information may be determined as the output respondent. In this way, the questioner can specify the respondent. Therefore, for example, a questioner who is likely to know an answer to a question to be asked can be designated in advance. Therefore, the possibility that an accurate answer to the question can be quickly obtained can be further increased.
[0013]
  Claims4As shown in FIG. 5, it is preferable to acquire the answer deadline for the question, and obtain and output the answer to the question until the obtained answer deadline. In this way, only answers acquired from the respondent by the reply deadline are output to the questioner. Therefore, for example, for a question that is meaningless if an answer is obtained after the answer deadline, if the questioner inputs the answer deadline in advance, no answer is output after the answer deadline has passed. Also, when the answer deadline for a question has passed, there is no need to wait for an answer to that question, reducing the processing load on the information system, and reducing the storage capacity for holding information related to questions and answers can do.
[0014]
  These natural language questions and answers can be obtained and output in a variety of data formats.5It is good to acquire as audio | voice data or text data as shown in. And claims6It is good to output as voice or text as shown in Claims7As shown in the above, it is preferable to enable conversion between voice data and text data. For example, audio data and text data may be converted to each other. In this way, for example, the question input by the voice from the questioner is output to the respondent by text display, the question input by the questioner in text is output to the respondent by voice, A text answer input from the respondent can be output to the questioner by voice, or a voice answer input from the respondent can be output as text. Note that the text data may be obtained in response to an input from a keyboard, or may be obtained by voice recognition by inputting as voice.
[0015]
  If the acquired question or answer contains an inappropriate keyword,8It is better not to output the question or answer as shown in. In this way, the system can be comfortably used without the questioner or the respondent receiving a message containing an inappropriate keyword.
[0016]
  By the way, the category storage means according to claim 1 stores a plurality of categories for classifying the question and the strength of each respondent in each category in association with each other. It may be stored based on information obtained from respondents, or claimed.9As shown in FIG. 4, it is preferable to acquire the evaluation of the questioner with respect to the answer and update the strength of the respondent based on the acquired evaluation. For example, when the questioner's evaluation is high, the skill level is increased, and when the questioner's evaluation is low, the skill level is decreased. In this way, since the level of proficiency changes according to the evaluation of the respondent, the respondent to be output can be determined more appropriately. This is particularly effective when the respondent himself sets an initial value of his / her strength.
[0017]
  Each information system described in each claim may be configured as an information system as a whole.10As shown in FIG. 1, one server (computer) may be provided with all means, for example, claims11As in the questioner-side terminal, the center, and the respondent-side terminal, each means and a portion in the means may be shared (distributed) by a plurality of devices such as computers. Claim10In the case of, the configuration is such that it is obtained from the network or output to the network.11In this case, it is configured to include a portion that is acquired from the user or output to the user. In addition, the information system may be configured by a plurality of different computers, installed in different places, and exchange of information between the computers may be performed by communication.
[0018]
  The questioner-side terminal and the respondent-side terminal of such an information system can be various terminals, but exhibit an excellent effect particularly when the terminal is usable from a mobile body. For example, claims12If the navigation device is used as shown in the figure, it is possible to easily obtain necessary information as an answer during movement..
[0019]
  by the way, MosquitoWhen extracting categories from questions in natural language, claims13As shown in the above, the terminal as the questioner side terminal extracts the category from the question in the natural language by voice, notifies the questioner of the extracted category, and confirms whether or not it is correct. When it is correct, it is preferable to provide a function of transmitting the extracted category and question. In this way, the processing load on the center side can be reduced, the questioner can easily determine the category, and the possibility of obtaining an accurate answer to the question can be increased.
[0020]
  In addition, it can be set as invention of a method by catching each means in the information system as described in each claim as a process step. That is, for example, the information system according to claim 1 is14It can be understood as a configuration for realizing the information acquisition method described in 1. And claims14According to the information acquisition method described in item 1, it is possible to obtain the same effect as that of the first aspect described above. Claim14As in claim 212By capturing each means in the information system described in each as a step, the information acquisition method can be invented, and according to the information acquisition method, an effect similar to the effect described above in the claims can be obtained. .
[0021]
  And claims15As shown in claims 1 to12When each means in the information system described in any of the above is realized by a computer system, for example, it can be provided as a program to be started on the computer system side. In the case of such a program, for example, the program is recorded on a computer-readable recording medium such as a flexible disk, a magneto-optical disk, a CD-ROM, a hard disk, a ROM, or a RAM, and is loaded into a computer system and started as necessary. It can also be used by loading and starting over a network.
[0022]
DETAILED DESCRIPTION OF THE INVENTION
Embodiments to which the present invention is applied will be described below with reference to the drawings. Needless to say, the embodiments of the present invention are not limited to the following examples, and can take various forms as long as they belong to the technical scope of the present invention.
[0023]
As shown in FIG. 1, the system includes a questioner-side terminal 10, a center 20, and an answerer-side terminal 30 that can be connected to the Internet.
The questioner-side terminal 10 and the respondent-side terminal 30 are computer systems that include a CPU, a memory such as a ROM / RAM, an I / O, and a bus line that connects them, and the I / O includes a microphone and a microphone. Audio input device including an A / D conversion device that converts the audio signal generated in step 1 into audio data, an audio output device that outputs audio from a speaker, etc., a display, a keyboard, and for connecting to the Internet for communication A communication device or the like is connected. Then, the CPU executes a program stored in the memory so that voice data is acquired by inputting voice from a microphone or the like, text data is recognized based on the acquired voice data, and text data is obtained as a recognition result. Voice recognition function for obtaining text, text acquisition function for obtaining text based on input to a keyboard, etc., transmission function for transmitting information obtained by these functions to the center 20, and reception function for receiving information transmitted from the center 20 By using these functions, the questioner side terminal 10 executes processing in the questioner side terminal described later, and the responder side terminal executes processing in the questioner side terminal described later.
[0024]
The questioner-side terminal 10 and the respondent-side terminal 30 can be configured as, for example, a navigation device, a mobile phone, a game machine, a personal computer, and other various information devices. In addition, you may comprise the questioner side terminal 10 and the respondent side terminal 30 as one terminal. In this case, this terminal functions as the questioner-side terminal 10 when the user makes a question or an evaluation to be described later (when it becomes a questioner), and when the user makes an answer (when it becomes an answerer), the respondent It functions as the side terminal 30.
[0025]
The center 20 is a computer that functions as a general server having a CPU, a memory such as a ROM / RAM, an I / O, and a bus line for connecting them, and the I / O is connected to a disk device and the Internet. Connected communication devices or the like. The disk device uses user information such as the user ID / name / address / age / gender / hobby / work content of the questioner and respondent, a list of question categories, and each category A good category table 40 (see FIG. 2A) showing the user's goodness degree and a list of inappropriate keywords are stored. The CPU executes a program stored in the memory, thereby realizing a function of receiving information from the questioner side terminal 10 and the answerer side terminal 30 via the Internet and transmitting information to these terminals. Using these functions, processing related to questions and answers, which will be described later, is performed. The center 20 can be connected to a large number of questioner-side terminals 10 and respondent-side terminals, and various processes are executed in parallel in a multitasking manner.
[0026]
When a questioner who is a user of the questioner-side terminal 10 asks a question, the content of the question is input to the questioner-side terminal 10 in a natural language from a voice or a keyboard. Further, the questioner selects a category corresponding to the content of the question to be input from a menu displayed on the questioner-side terminal 10. For example, as shown in FIG. 2B, this category is included in a broad category such as “(1) Travel / tourism, (2) Food and drink, (3) Hobby ...”, and the broad category. It consists of a hierarchical structure consisting of small categories such as “parks, theme parks, temples and shrines”. A list of categories as shown in FIG. 2B showing the classification of the questions, that is, the genre (category) of the questions, is obtained in advance by the questioner-side terminal 10 from the center 20 and stored in the menu. It is displayed in the form and is selected using a keyboard or the like. Further, the questioner can input position information that is a place related to the question, an answer deadline for specifying a time limit of the answer, and the like as necessary.
[0027]
For example, as shown in FIG. 1, a questioner inputs a voice question “Do you have a delicious and rich pork bone ramen shop in Nagoya?”, And the category of this question is “Food & Drink> Ramen” When “Nagoya” is selected as the place and “30 minutes” is selected as the response time limit, the questioner-side terminal 10 transmits the inquiry information together with the user ID to the center. Note that the question category, position information, answer deadline, and the like may be obtained from the result of speech recognition of the content of natural language. The question is acquired, for example, as text data or voice data, and transmitted to the center 20 as mail or voice mail.
[0028]
The center 20 receives the inquiry information and the user ID transmitted from the questioner-side terminal 10 and receives the inquiry using the good category table 40 or the like stored in the center 20 as shown in FIG. Determine the respondents for whom information is to be sent. As shown in FIG. 2A, the good category table 40 stores the goodness of each category for each user. Then, the inquiry information is transmitted to the respondent side terminal 30 of the determined respondent to be transmitted.
[0029]
The respondent side terminal 30 receives this inquiry information, and notifies a natural language question included in the received inquiry information as a voice or a text. When the respondent answers the informed question in this way, for example, as shown in FIG. 1, “There is a ramen shop called“ A store ”” as a natural voice or text. Enter in words. The respondent side terminal 30 transmits the input natural language speech data or text data answer to the center 20 as mail or voice mail together with the user ID of the respondent.
[0030]
When the center 20 receives the answer and the user ID from the answerer side terminal 30, the center 20 transmits the answer and the user ID of the answerer to the questioner side terminal 10.
When the questioner-side terminal 10 receives the answer and the user ID of the answerer, the answerer side 10 notifies the answer by voice or text. As a result, the user can obtain an answer to his / her question. When the questioner inputs an evaluation for the informed answer to the questioner-side terminal 10, the questioner-side terminal 10 transmits the evaluation to the center 20, and the center 20 receives the evaluation. Based on the evaluation, the good category table 40 described above is updated.
[0031]
Hereinafter, the contents of the processing of each device will be described with reference to FIGS. 3 to 11 for more detailed examples of the processing described above.
FIG. 3 is a flowchart showing a flow of question-time processing executed by the questioner-side terminal 10 when the question input instruction is a keyboard or voice input in the questioner-side terminal 10. FIG. 4 is a functional block diagram showing the contents of the question processing in FIG.
[0032]
As shown in FIG. 3, the questioner side terminal 10 first sets the speech recognition dictionary as a dictionary for natural language input to enable input of inquiry information by voice, and also allows manual input (keyboard and button input, screen display). It is possible to input inquiry information by operation (S110). And the input of the inquiry information of the natural language by the voice or the text from the user is accepted (S120) (the function to perform this voice or manual input corresponds to the user interface shown in FIG. 4). The input contents by voice or manual input can include a question category, a place (position information), a respondent, an answer deadline, etc. in addition to the question. For example, after inputting the contents of the question, “category is XX, location is Aichi prefecture, respondent is XX, and reply deadline is within 30 minutes” can be input. The question category, location (position information), respondent, answer deadline, and the like may be included in the question. For example, the location (location information) may be included in the question, such as “Is there a delicious ramen shop in Nagoya?” Or “Is there a beautiful place for autumn leaves?” Further, the reply deadline may be designated by a relative time from the present such as “within 1 hour” or may be designated by an absolute time such as “until 17:30”. When the input content input in this way is input by voice, the voice data and text data obtained by voice recognition of the voice data are stored as inquiry information, and when input by manual input, And stored as inquiry information as text data (S130, corresponding to the inquiry record in FIG. 4).
[0033]
In subsequent S140, it is determined based on the text data stored in S130 whether or not the respondent is specified as the inquiry information. If the respondent is specified (S140: YES), the process proceeds to S200. If it is not designated (S140: NO), the process proceeds to S150.
[0034]
In S150, it is determined based on the text data stored in S130 whether or not clear position information (position information previously stored in the memory as a list of position information) is designated as inquiry information. If information is specified (S150: YES), the process proceeds to S170. If position information is not specified (S150: NO), the process proceeds to S160. For example, if the text data of the inquiry information that does not include clear position information such as “Is there a place where autumn leaves are beautiful in this area?” Is stored, or if the position information is not included, the process proceeds to S160. To do.
[0035]
In S160, the current position is detected by the GPS provided in the interrogator side terminal 10, and information on the area corresponding to the detected current position is added to the inquiry information and stored. For example, “Place is Aichi Prefecture” is added.
In S170, if the position information is included in the text data of the inquiry information, the position information is extracted, and the extracted position information is stored as inquiry information for transmission in addition to the inquiry information stored in the memory in S130 ( This corresponds to the position information generation in FIG. For example, if the text data of the inquiry information is “Is there a delicious ramen shop in Nagoya?”, “Nagoya” is extracted and stored as location information.
[0036]
In S180, if the text data of the inquiry information includes a category, the category is extracted and the category database stored in the memory shown in FIG. 4 is referred to.
In S190, it is determined by referring to the category database storing the category list whether the category extracted in S180 is a category in the category list. If the category is in the category database, the category is transmitted. Store in addition to the inquiry information. If the category could not be extracted, or if the extracted category is a category that is not in the category database, the category stored in the category database is displayed on the display device in a menu format, and selected by voice or manual input. The selected category is stored as inquiry information for transmission. The category database stores categories in the hierarchical structure shown in FIG. 2B, and the menu is displayed based on this hierarchical structure. Therefore, the user (questioner) can determine the genre (category) of the inquiry by simply following the category hierarchy. A plurality of categories may be designated.
[0037]
On the other hand, in S200, the user information constructed from the past communication history is referred to, and in the subsequent S210, the user ID of the respondent designated as the inquiry information is acquired and added to the inquiry information for transmission ("" in FIG. 4). It corresponds to “user information” and “user ID”.) Note that the questioner may directly input the user ID of the respondent.
[0038]
In S220, it is determined whether or not the answer deadline is included in the text data of the inquiry information. If included (S230: YES), the process proceeds to S240, and if not included (S230: NO). , The process proceeds to S230.
In S230, the initial setting time stored in advance in the memory is set as the inquiry information reply time limit. The initial setting time can be, for example, 1 hour, 6 hours, 1 day from the present time.
[0039]
In S240, the response time limit is acquired from the inquiry information and added to the inquiry information for transmission (corresponding to the generation of the response time limit in FIG. 4). If the response deadline is specified as a relative time from the current time, such as “within 30 minutes”, it is converted to an absolute time (date, time, etc.) and sent as the response deadline. To the inquiry information.
[0040]
In S250, the user information of the questioner is referred to, and the user ID of the questioner is acquired in S260.
In S270, the inquiry information for transmission and the user ID of the questioner acquired in S260 are transmitted to the center 20.
[0041]
In S280, it waits for the response information from the center and the user ID of the respondent to be transmitted, and receives the response information and the user ID of the respondent (corresponding to the acquisition of the response information in FIG. 4). Then, the answer information and the user ID of the answerer are stored in the memory (S290). If the answer included in the answer information is text data, the answer is displayed on the display. If the answer is voice data, the speaker is used as a voice. Etc. (S300). Note that when the answer is text data, a voice corresponding to the text data may be synthesized and output.
[0042]
In S310, a display prompting the user to select whether or not to input an evaluation for the answer is displayed. When an instruction to evaluate the answer is input from a keyboard or the like (S310: YES), the process proceeds to S320 in FIG. If the instruction to perform the evaluation is not input (S310: NO), the process proceeds to S380 in FIG.
[0043]
In S320, voice input and manual input (keyboard or button input, screen operation, etc.) are enabled. In S330, a natural language input by voice and text from the user is accepted, and in the same manner as in S120, at least voice data or text data is received. Get either one. For example, the questioner inputs an evaluation such as “OK” or “OK”, and obtains the voice data and text data of the voice recognition result.
[0044]
In S340, an evaluation value is determined based on the text data obtained in S330. For example, the evaluation value is a five-level value from 1 to 5. For example, evaluation value 3 corresponds to evaluation value 3 when “good” is included, and evaluation value 4 corresponds to evaluation value 4 when “good” is included. An evaluation value is determined based on data such as an evaluation value determination table stored in addition (corresponding to evaluation value generation in FIG. 4). Of course, the evaluation value may be directly input by voice or manual input.
[0045]
In subsequent S350, the user information of the questioner stored in the memory is referred to, and the user ID of the questioner is acquired in S360.
In S370, the evaluation value determined in S340, the user ID of the respondent, and the category included in the answer information received in S280 are transmitted as evaluation information to the center 20 together with the user ID of the questioner acquired in S360.
[0046]
In S380, information prompting to input whether or not to end the question is displayed. When an instruction to end is input (S380: YES), this process ends and an instruction to continue is input. (S380: NO), the process proceeds to S110 in FIG. When an instruction to continue is input, for example, a message asking the questioner whether or not to specify the respondent of this answer as the answerer of the next question is displayed. When an instruction to be specified is input, the respondent may be included in the inquiry information. In this way, it is possible to easily send questions continuously to respondents who have provided useful answers.
[0047]
Next, processing of the center 20 when the center 20 receives a question (when a question is received) will be described with reference to the flowchart of FIG. FIG. 7 shows a functional block diagram of the center 20.
In S410 of FIG. 6, the inquiry information including the question (inquiry), category, position information, answer deadline, answerer ID, etc. transmitted from the questioner side terminal 10 and the user ID of the questioner are received.
[0048]
In S420, user information stored in the center 20 is acquired. In S430, it is checked whether the user ID received in S410 is included in the user information acquired in S420. In S440, it is determined whether or not there is corresponding user information. If there is corresponding user information (S440: YES), the process proceeds to S450, but if there is no corresponding user information (S440). : NO), the process proceeds to S480. In S480, an error message is transmitted to the interrogator terminal 10 and the inquiry information is deleted. That is, the user of the questioner is authenticated, and only the processing after S450 is performed (corresponding to “user check and acquisition of user information from user ID” in FIG. 7).
[0049]
In S450, the inappropriate keyword list stored in the center 20 is referred to, and in S460, it is checked whether or not an inappropriate keyword is included in the inquiry information. When only voice data can be extracted from the inquiry information, the voice data may be recognized by voice recognition to determine the presence or absence of an inappropriate keyword.
[0050]
In S470, it is determined whether or not the content of the inquiry information is appropriate from the result of the check in S460. If the contents of the inquiry information are appropriate (S470: YES), the process proceeds to S490. If the contents of the inquiry information is not appropriate (S470: NO), the process proceeds to S480 (for checking the question in FIG. 7). Equivalent to). In S480, an error message is transmitted to the terminal, and the inquiry information is deleted.
[0051]
In S490, the response time limit included in the inquiry information is referred to and stored in S500 (corresponding to the storage of the response time limit in FIG. 7).
In S510, it is determined whether or not an answerer is specified based on whether or not the inquiry information includes an answerer ID. If an answerer is specified (S510: YES), the process proceeds to S550. When the respondent is not designated (S510: NO), the process proceeds to S520.
[0052]
In S520, the location information included in the inquiry information and the user information stored in the center 20 are referred to. In S530, the category included in the inquiry information and the specialty category table 40 stored in the center 20 are referred to. The respondents are selected (corresponding to “selecting respondents from the inquiry category and location information and the good category table” in FIG. 7). As shown in FIG. This stores the strength of each category for each (respondent). That is, what genre (category) the user is good at (is familiar with) is stored as the goodness, and in FIG. 2A, the goodness is expressed numerically. FIG. 2A shows that the larger the numerical value, the better, and the smaller the numerical value, the poorer. Note that the level of goodness may be expressed by, for example, a vector instead of a numerical value. This goodness level may be input by the user (respondent) himself, for example, at the time of user registration, or when the user information is registered, such as age, gender, hobby, work content, etc. included in the user information. It may be automatically determined from the profile.
[0053]
In selecting the respondent in S540, for example, a user whose location specified as the position information is included in the address stored as the user information is extracted. For example, when “Nagoya City” is included as position information, a user whose address stored as user information includes “Nagoya City” is extracted. Then, referring to the goodness of the category included in the inquiry information for the extracted user, for example, a predetermined number of people (for example, 10 people) in descending order of the goodness of the category, the answers to which the inquiry information is output As a person. It should be noted that there may be a plurality of respondents as described above, or only one responder, such as a person who has the highest level of proficiency.
[0054]
In S550, the user information of the respondent is referred to, the user information is acquired in S560, and the inquiry information in S570 and the user ID of the questioner received in S410 are transmitted to the respondent's answerer side terminal 30.
Next, the processing in the respondent side terminal 30 when the inquiry information transmitted from the center 20 and the user ID of the questioner are received will be described with reference to the flowchart of FIG. 8 and the functional block diagram of FIG. .
[0055]
In S610 of FIG. 8, the inquiry information and the user ID of the questioner are received from the center 20, stored in the memory in S620, and the question in the inquiry information is output to the respondent by voice or display in S630. That is, when the question is text data, the question is displayed on the display, and when the question is sound data, the sound is output from a speaker or the like. Note that when the question is text data, a voice corresponding to the text data may be synthesized and output.
[0056]
In S640, a display prompting the user to select whether or not to input an answer to the output question is displayed. If an instruction to evaluate the answer is input manually or by voice (S640: YES), S650 is displayed. When an instruction not to evaluate is input (S640: NO), the process proceeds to S700 and a non-response message is transmitted to the center 20.
[0057]
In S650, the speech recognition dictionary is set as a dictionary for natural language input to enable input of inquiry information by voice, and input of inquiry information by manual input (keyboard, button input, screen operation, etc.) is enabled. . In step S660, an input of a natural language answer by voice or text from the user is accepted, and the answer is acquired as voice data or text data. The function of performing voice or manual input corresponds to the user interface shown in FIG.
[0058]
In S670, the user information of the respondent stored in the memory is referred to, and the user ID of the respondent is acquired in S680 (corresponding to the user ID acquisition in FIG. 9).
In step S690, the answer, the user ID of the questioner received in step S610, and the inquiry information are transmitted as answer information to the center 20 together with the user ID of the respondent (corresponding to the communication in FIG. 9).
[0059]
Next, processing in the center 20 when receiving the reply information and the respondent's user ID transmitted from the respondent side terminal 30 will be described with reference to the flowchart of FIG. 10 and the functional block diagram of FIG. .
In S810 of FIG. 10, the response information and the user ID of the respondent are received. In step S820, the user information is referred to. In step S830, it is checked whether the user ID received in step S810 is included in the user information referred to in step S820. In S840, it is determined whether or not there is corresponding user information. If there is corresponding user information (S840: YES), the process proceeds to S850, and if there is no corresponding user information (S840: NO), the process proceeds to S910. That is, the user authentication of the respondent is performed (corresponding to “user check and user information acquisition from user ID” in FIG. 7).
[0060]
In S850, the reply deadline included in the inquiry information in the reply information received in S810 is referred to, and it is determined in S860 based on the reply deadline and the current date and time whether or not a reply has been made within this reply deadline ( This corresponds to the check of the response deadline time in FIG. If the reply is made within the reply deadline (S860: YES), the process proceeds to S870. If the reply is not made within the reply deadline (S860: NO), the process proceeds to S910.
[0061]
In S870 to S890, as in S450 to S470 of FIG. 6 described above, the inappropriate keyword list stored in the center 20 is referred to (S870), and an inappropriate keyword is not included in the answer to the inquiry information. Is checked (S880). Then, it is determined from the check result whether the content of the inquiry information is appropriate (S890). If the content of the inquiry information is appropriate (S890: YES), the process proceeds to S900. If the content of the inquiry information is not appropriate (S890: NO), the process proceeds to S910 (for checking the answer in FIG. 7). Equivalent to).
[0062]
In S900, the answer information and the user ID of the respondent are transmitted to the questioner-side terminal 10 based on the user ID of the questioner included in the answer information. In this way, answer information including the answer of the respondent can be sent to the questioner-side terminal 10. In the questioner-side terminal 10, the answer is notified to the respondent by the processing shown in S280 and subsequent steps in FIG. 3, and the evaluation information is transmitted to the center 20 by the processing in FIG. 5 as necessary.
[0063]
On the other hand, in S910, an error message is transmitted to the respondent side terminal 30, and the answer is deleted in S910.
Next, the processing of the center 20 when receiving evaluation information from the questioner-side terminal 10 will be described with reference to the flowchart of FIG. 11 and the functional block diagram of FIG.
[0064]
In S1010 of FIG. 11, the evaluation information transmitted from the questioner side terminal 10 and the user ID of the questioner are received.
In S1020, the user information stored in the center 20 is acquired. In S1030, it is checked whether the user ID of the questioner received in S1010 is included in the user information acquired in S1020. In S1040, it is determined whether or not there is corresponding user information. If there is corresponding user information (S1040: YES), the process proceeds to S1050, but if there is no corresponding user information (S1040). : NO), the process proceeds to S1070. In S1070, an error message is transmitted to the questioner side terminal 10. That is, the user authentication of the questioner is performed, and the processing after S1050 is performed only when authenticated (corresponding to “user check and user information acquisition from user ID” in FIG. 7).
[0065]
In S1050, the specialty category table 40 is referred to. In S1060, the user ID of the respondent included in the evaluation information received in S1010 in the specialty category table 40 and the strength corresponding to the category are included in the evaluation information. Update based on the evaluation (corresponding to update of the specialty category table in FIG. 7). For example, if the evaluation value is 4, one tenth of 0.4 is added to the intention.
[0066]
With the configuration as described above, it is possible to select an answerer who is good in the field and obtain an answer to an ambiguous inquiry of a questioner who is difficult to understand by the system.
In the above embodiment, the respondents are determined based on the category, position information, etc. For example, the users may be grouped in advance and the respondents to be inquired may be determined within the range of the group. Good. Further, only respondents in the group designated by the user may be determined as inquired respondents. That is, a group may be designated as additional information in the claims. The group may be set so that the respondent can set his / her own group, or may be automatically set from user information stored in the center 20. For example, user information may be referred to and grouped by age or gender.
[0067]
In addition, as a method of sending inquiry information to the respondents narrowed down based on the category, respondent, location (location information), etc. as the inquiry destination, it may be sent to all the respondents who have narrowed down, You may send a limited number of respondents, or send at least one of the narrowest respondents from the top category value and wait for feedback from the questioner. It may be decided whether to send to the next respondent.
[0068]
In the above embodiment, the center 20 updates the good category table 40 based on the evaluation information when the evaluation information is received from the questioner side terminal 10. For example, the center 20 updates the answer information from the responder side terminal 30. When it is received, it may be updated so as to improve the evaluation. In addition, when the reply information is not transmitted before the reply deadline, it may be updated so as to lower the evaluation.
[0069]
  In this embodiment, the claims10When the invention described in each claim is captured with the configuration shown in FIG. 5, the center 20 corresponds to an information system that is a server, the disk device that stores the good category table 40 corresponds to a category goodness storage means, and user information is stored. The stored disk device corresponds to the respondent information storage means, and the processing of S410 corresponds to the processing as question acquisition means, category acquisition means, additional information acquisition means, answer deadline acquisition means, S450 to 470, S870 to 890 The processing corresponds to processing as an inappropriate keyword determination unit, the processing in S510 to 560 corresponds to processing as an answerer determination unit, and the processing in S570 corresponds to processing as a question output unit. Further, the process of S810 corresponds to a process as an answer acquisition unit, and the process of S900 corresponds to a process as an answer output unit. The process of S1010 corresponds to the process as the evaluation acquisition unit, and the process of S1060 corresponds to the process as the strength update unit.
[0070]
  Claims11When the invention described in each claim is captured with the configuration shown in FIG. 4, the questioner-side terminal 10 corresponds to the questioner-side terminal, the center 20 corresponds to the center, and the respondent-side terminal 30 corresponds to the respondent-side terminal. And the system comprised from the questioner side terminal 10, the center 20, and the respondent side terminal 30 shown in FIGS. 1-11 corresponds to an information system.
[Brief description of the drawings]
FIG. 1 is an explanatory diagram illustrating a configuration and an example of use of an information system.
FIG. 2 is an explanatory diagram showing a good category table and a list of categories.
FIG. 3 is a flowchart showing processing at the time of a question in the questioner side terminal;
FIG. 4 is a functional block diagram of a questioner side terminal;
FIG. 5 is a flowchart showing processing at the time of evaluation in the questioner-side terminal.
FIG. 6 is a flowchart showing processing at the time of question reception at the center.
FIG. 7 is a functional block diagram of the center.
FIG. 8 is a flowchart showing processing in the respondent side terminal.
FIG. 9 is a functional block diagram of the respondent side terminal.
FIG. 10 is a flowchart showing processing at the time of answer reception at the center.
FIG. 11 is a flowchart showing processing at the time of evaluation reception at the center.
[Explanation of symbols]
10 ... Questioner side terminal
20 ... Center
30 ... Respondent side terminal
40 ... Professional category table

Claims (15)

A plurality of categories for classifying questions, and a category goodness storage means for storing the goodness of each respondent in each category in association with each other;
Respondent information storage means for storing respondent information including position information indicating at least the area where the respondent is good at the respondent;
A question acquisition means for acquiring questions in natural language from the questioner;
Category acquisition means for acquiring the category of the question in the natural language;
Question position information acquisition means for acquiring position information related to the question from the questioner;
Additional information acquisition means for acquiring additional information from the questioner for the question including at least the position information acquired by the question position information acquisition means;
Respondent candidate determination means for determining respondent candidates based on the respondent information stored by the respondent information storage means and the additional information acquired by the additional information acquisition means;
The answer in the category extracting good degree good degree from good degree of each stored respondents in the storage means the respondent candidate determined by the respondent candidate determining means, the category obtained by said category obtaining unit candidate for on the basis of the good degree performs ranking on the respondent candidate, and respondents determining means for determining a respondent's output target questions at least upper one person obtained by the interrogator acquiring means,
Respondent storage means for storing the respondent information stored by the respondent information storage means regarding the determined respondent;
Question output means for outputting the question to the respondent stored by the respondent storage means;
An answer acquisition means for acquiring an answer in natural language to the question;
An information system comprising: an answer output means for outputting the answer acquired by the answer acquisition means to the questioner,
The question position information acquisition means acquires information indicating the current position of the questioner as position information related to the question when the questioner does not explicitly specify position information related to the question. Characteristic information system. The information system according to claim 1,
The question position information acquisition means acquires information representing a real-time current position using a GPS provided in a questioner's terminal or the like . The information system according to claim 1 or 2,
As the additional information, including respondent designation information for designating a respondent,
The respondent determination means determines the respondent designated by the respondent designation information as the respondent to be output . In the information system in any one of Claims 1-3 ,
It has an answer deadline acquisition means to get the answer deadline for questions from the questioner,
The answer acquisition means acquires an answer to the question until the answer deadline acquired by the answer deadline acquisition means,
The answer output means outputs an answer to the question until the answer deadline acquired by the answer deadline acquiring means . In the information system according to any one of claims 1 to 4 ,
The question acquisition means and the answer acquisition means acquire the question or the answer as voice data or text data . In the information system according to any one of claims 1 to 5 ,
The question output means and the answer output means output the question or the answer as voice or text . The information system according to claim 5 or 6 ,
An information system comprising data conversion means for converting a data format between voice data and text data . In the information system in any one of Claims 1-7,
An inappropriate keyword determining means for determining whether or not an inappropriate keyword is included in the question or the answer,
The information system according to claim 1 , wherein the inappropriate keyword determination unit prohibits the output of the question or the answer when it is determined that the inappropriate keyword is included . In the information system in any one of Claims 1-8,
Evaluation acquisition means for acquiring a questioner's evaluation for the answer output by the answer output means;
An information system comprising: a goodness degree updating means for updating the goodness of the respondent stored in the category storage means based on the evaluation acquired by the evaluation acquisition means . In the information system in any one of Claims 1-9,
The information system is a server, and the server is configured to be able to communicate with the interrogator side terminal and the respondent side terminal via the network.
The question acquisition means receives the question from the questioner side terminal,
The category acquisition means receives the category from the questioner side terminal,
The question output means transmits the question to the respondent side terminal,
The answer acquisition means receives the answer from the respondent side terminal,
The answer output means transmits the answer to the questioner side terminal . In the information system in any one of Claims 1-9 ,
The information system includes a questioner side terminal, a center, and an answerer side terminal,
The center is configured to be able to communicate with the interrogator side terminal and the respondent side terminal via a network,
The question acquisition means is composed of the questioner-side terminal that inputs a question from a questioner and transmits it to the center, and the center that receives a question transmitted from the questioner-side terminal,
The category acquisition means includes the questioner side terminal that transmits the category acquired from the questioner to the center, and the center that receives the question transmitted from the questioner side terminal,
The question output means includes the center that transmits the question to the respondent side terminal, and a respondent side terminal that receives the question transmitted from the center and outputs the question to the respondent.
The answer acquisition means includes the answerer side terminal that transmits the answer to the center, and a center that receives an answer transmitted from the answerer side terminal,
The answer output means includes the center that transmits the answer to the questioner-side terminal and the answerer-side terminal that receives the answer transmitted from the center . The information system according to claim 10 or 11 ,
At least one of the questioner side terminal and the answerer side terminal is a navigation device . A terminal as the interrogator side terminal in the information system according to any one of claims 10 to 12,
A voice question inputted from the questioner is recognized, and based on the recognition result, a category corresponding to the voice inputted from a list of categories stored in advance is determined and notified, and the notified category is the questioner. A terminal that transmits an informed category and a question input from the questioner when an instruction indicating that the category is appropriate as a question category by voice input from the questioner is input . The category proficiency storage means stores a plurality of categories for classifying questions and the proficiency for each respondent in each category in association with each other, and the respondent information storage means stores at least position information indicating a region where the respondent is good at Including the respondent information, the question acquisition means acquires a question in natural language from the questioner, the category acquisition means acquires the category of the question in natural language, and the question position information acquisition means receives the question from the questioner. The position information related to the question is acquired, and the additional information acquisition means acquires additional information from the questioner for the question including at least the position information acquired by the question position information acquisition means , and respondent candidate determination means Is based on the respondent information stored by the respondent information storage means and the additional information acquired by the additional information acquisition means. The respondent determination means extracts the goodness of the respondent candidate determined by the respondent candidate determination means from the goodness of each respondent stored in the category goodness storage means, and the category acquisition means Ranking the respondent candidates based on the proficiency of the candidate respondents in the category acquired by determining at least the top one as respondents to be output of the questions acquired by the question acquiring means; The respondent storage means stores the respondent information stored by the respondent information storage means regarding the determined respondent, and the question output means asks the question to the respondent stored by the respondent storage means. output, the answer acquiring unit acquires an answer in natural language for the question, the answer output means has been responded acquired by the answer acquiring unit to the questioner An information acquisition method of and outputs, wherein when the question positional information acquiring unit that does not explicitly specify the position information the questioner is related to the question, the question information indicating the current position of the questioner An information acquisition method, characterized in that the position information is related to . The program for functioning a computer as an information system in any one of Claims 1-12.

Images