JP2011060218A - Reply system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide information that is appropriate to a user's actual knowledge level even when assuming that the user inputs his/her already known keyword, or the like. <P>SOLUTION: A reply device receives the input of a first keyword about a question from the user, calculates a weighted average value between the first keyword whose input is received and a second keyword whose input is received from the user during the past predetermined period, sets the calculated weighted average value as the present knowledge level of the user, and determines and outputs a reply corresponding to the set knowledge level from a means which stores a plurality of kinds of replies in accordance with the knowledge level for each question. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a technique for providing an answer that matches a questioner's knowledge level.

  In recent years, manufacturers of products and the like often provide support services that accept inquiries from users via the Internet and automatically provide answers to the inquiries.

  For example, in Patent Document 1 below, based on a search keyword received from a user who has a problem, the user's intention and purpose are specified, an inquiry for specifying the cause of the problem, and a response from the user to the inquiry In addition, a technique for presenting a solution that the user wants to obtain by repeatedly grasping the user's browsing situation with respect to the annotation presented during the interview is disclosed.

JP 2008-21101 A

  In the technique of Patent Document 1, when the user selects an annotation presented during an inquiry by clicking or the like, the knowledge level (user level value) of the user is updated based on the level value corresponding to the annotation. The

  Specifically, in the technique described in Patent Document 1, when the user selects an annotation with a low level value, the user level value is updated so as to decrease, while the user selects an annotation with a high level value. If so, the user level value is not updated.

  This is because the technique described in Patent Document 1 is based on the premise that an annotation is selected in order to obtain an explanation of an unknown term or the like. Therefore, it is not possible to select an annotation with a level value higher than the current user level value. This is because it is determined as a natural action for the user at the user level value.

  However, such a technique described in Patent Document 1 does not operate properly in a case where a query is input by inputting a keyword that the user already knows. That is, in the case where the user makes an inquiry by entering a keyword, the fact that the user has entered a high level keyword means that the knowledge level of the user is high. This is because the knowledge level of the user is not updated even in such a case. As a result, there is a risk that information such as explanatory text that is not familiar with the actual knowledge level is presented.

  In view of this problem, an object of the present invention is to provide a new framework that can provide information according to the actual knowledge level of the user even when the user already inputs a keyword or the like.

  The answering device of the present invention is a weighting means for receiving an input of a first keyword related to a question from a user, a first keyword that has received the input, and a second keyword that has received an input from the user in a past predetermined period. From the means for obtaining a weighted average value and setting the obtained weighted average value as the current knowledge level of the user, and means for storing a plurality of types of answers according to the knowledge level for each question. Means for determining and outputting an answer corresponding to.

  In an embodiment of the present invention, the means for storing the answer includes means for storing a set of a plurality of questions and answers, and means for storing information for supplementing the answer in association with the knowledge level, The means for outputting stores means for acquiring an answer to a question related to the first keyword that has received the input from means for storing the set of the plurality of questions and answers, and information for supplementing the answer Means for acquiring supplementary information corresponding to the set knowledge level, and using the acquired supplementary information, the content of the acquired answer is updated and determined as an answer corresponding to the set knowledge level It is preferable to comprise so as to include a means to perform.

  In another embodiment of the present invention, the means for setting the knowledge level can be configured to calculate the weighted average value using the following equation:

  In the above formula, the old weighted average value is the weighted average value calculated for the user last time, the number of search terms (cumulative) is the number of second keywords that have received the input, and the number of search terms (new) is the input The number of first keywords that have received Σ and the Σ weight can be the sum of the weights of the first keywords.

  An answer system of the present invention is an answer system including a user terminal and an answer device configured to be able to communicate with the user terminal, wherein the user terminal receives a first keyword related to a question input by a user. Means for transmitting to an answering device, the answering device accepting an input of the first keyword from the user terminal, an input of the first keyword accepting the input, and an input from the user in a past predetermined period. Means for determining a weighted average value with the second keyword, setting the determined weighted average value as the current knowledge level of the user, and storing a plurality of types of answers for each question according to the knowledge level Means for determining and outputting an answer corresponding to the set knowledge level from the means.

  The answering method of the present invention includes a step of receiving an input of a first keyword related to a question from a user, a weight of the first keyword that has received the input, and a second keyword that has received an input from the user in a past predetermined period. A set average level, a step of setting the calculated weighted average value as the current knowledge level of the user, and a means for storing a plurality of types of answers according to the knowledge level for each question. And determining and outputting an answer corresponding to.

  The computer program of the present invention is a computer program for causing an information processing apparatus to execute the answering method.

  In the present invention, the term “means” does not simply mean a physical means, but includes a case where the functions of the means are realized by software. Also, the functions of one means or device may be realized by two or more physical means or devices, or the functions of two or more means or devices may be realized by one physical means or device.

  According to the present invention, the weighted average value of the keyword newly received from the user (first keyword) and the keyword (second keyword) received from the user within the past predetermined period, By setting the current knowledge level of the user, if the weight of the first keyword is higher than the weighted average value of the second keyword, the weight is set higher, and the weight of the first keyword is set to the weight of the second keyword. If it is lower than the average value, it is set lower. Therefore, even when the user newly inputs a keyword he / she knows, the user's current knowledge level can be set more accurately, and an answer corresponding to the set knowledge level can be output.

It is a figure which shows the whole structure of the reply system of 1st Embodiment. It is a figure which shows the example of the data structure of a user information storage means. It is a figure which shows the example of the data structure of a classification | category memory | storage means. It is a figure which shows the example of the data structure of a weight memory | storage means. It is a figure which shows the example of the data structure of a question answer memory | storage means. It is a figure which shows the example of the data structure of a supplementary information storage means. It is a figure which shows the example of the data structure of a level reference | standard storage means. It is a figure which shows the example of the data structure of a reply temporary storage means. It is a figure which shows the state by which the reply stored in the reply temporary storage means was updated. It is a sequence diagram which shows the flow of the reply process in a user terminal and an answer apparatus. It is a figure which shows the example of an answer search screen. It is a figure which shows the state by which the data of the user information storage means were updated. It is a figure which shows the example of the question and answer which are displayed on the display of a user terminal. It is a figure which shows another example of the question and answer which are displayed on the display of a user terminal. It is a figure which shows the example of the email containing question / answer search information.

  Embodiments of the present invention will be described below with reference to the drawings.

  FIG. 1 shows the overall configuration of the answer system 1 of the first embodiment.

  The answer system 1 is a system for supporting a user of a certain product, for example, and automatically determines and provides an answer corresponding to a user's knowledge level for a keyword related to a question received from the user. In the present invention, the knowledge level is a value indicating the degree of knowledge and the degree of some skill.

  In order to realize such a framework, the answer system 1 includes a user terminal 3 operated by a user and an answer device 5 configured to be communicable with the user terminal 3.

  The user terminal 3 is configured by installing a computer program such as browser software on a general personal computer, and is input by the user as a means that is functionally realized by the CPU reading the program onto the memory and executing it. The information input / output means 7 for transmitting information (question / answer search information) for searching for a question and an answer to the answering device 5 is provided. The question / answer search information includes, for example, a keyword (first keyword) newly input by a user for searching for questions and answers, a classification indicating a technical field to which the first keyword belongs (computer, car, etc.) ) Or when the question and answer to be searched are stored (updated) in the answering device 5 (that is, when a certain period is specified, only the question and answer stored within that period are searched) Such information is included. The user terminal 3 is used in a state where input devices such as a keyboard and a mouse and output devices such as a display are connected.

  The answering device 5 is configured by installing the computer program according to the first embodiment on a general server computer, and is functionally realized by the CPU reading the program on the memory and executing it. 9 and a storage unit 11 functionally realized using a storage device such as a hard disk.

  The processing unit 9 includes a keyword receiving unit 13, a level acquiring unit 15, a weight acquiring unit 17, a level determining unit 19, and an answer output unit. The answer output means includes an answer acquisition means 21, a supplementary information acquisition means 23, and an answer determination means 25.

  The storage unit 11 includes user information storage means 31, classification storage means 33, weight storage means 35, answer storage means, level reference storage means 37, and answer temporary storage means 39. The answer storage means includes question answer storage means 41 and supplemental information storage means 43.

  The user information storage unit 31 stores the knowledge level for each user and the number of keywords (second keywords) received in the past. Specifically, as shown in FIG. 2, the user ID (UID), the knowledge level of the user corresponding to the UID, and the number of second keywords received by the keyword receiving means 13 from the user (search The number of terms) is stored in association with each other.

  The classification storage unit 33 stores a plurality of types of classifications. Specifically, as shown in FIG. 3, a classification ID (CID) and a classification are stored in association with each other.

  The weight storage means 35 stores the weight for each keyword. Specifically, as shown in FIG. 4, a classification ID (CID), a keyword (search term), and a weight of the keyword are stored in association with each other.

  The answer storage means stores a plurality of types of answers according to the knowledge level for each question by the cooperation of the question answer storage means 41 and the supplementary information storage means 43 as described below.

  The question answer storage means 41 stores a set of a plurality of questions and answers. Specifically, as shown in FIG. 5, a question ID (QID), a classification ID (CID), a question, an answer to the question, and a time when the question and the answer are stored (or the answer is updated). Stored) in association with each other.

  The supplementary information storage unit 43 stores information for supplementing the answer in association with the knowledge level. Specifically, as shown in FIG. 6, a supplementary description is added to the term as a classification ID (CID), a term to be supplemented included in the answer, and information for supplementing the term. A plurality of types of term supplementary information are stored in association with each other. In addition, multiple types of supplementary terminology information are prepared for each level point that is predetermined according to the knowledge level so that answers with different contents depending on the knowledge level (answers including explanations with different difficulty levels) are presented. Has been.

  As shown in FIG. 7, the level reference storage means 37 stores a level point, the lowest point of the knowledge level belonging to the level point, and the highest point of the knowledge level belonging to the level point in association with each other.

  The answer temporary storage means 39 stores the answer acquired by the answer acquisition means 21 from the question answer storage means 41 or stores an updated answer performed using supplementary information. Specifically, when the input of the first keyword is received from a certain user, as shown in FIGS. 8 and 9, the UID of the user, the CID of the classification to which the first keyword belongs, and the first The QID of the question determined based on the keyword is stored in association with the answer corresponding to the question.

  The keyword accepting unit 13 accepts input of question / answer search information including the first keyword from the user terminal 3.

  The level acquisition unit 15 acquires the knowledge level and the number of second keywords (the number of search terms shown in FIG. 2) from the user information storage unit 31 for the user who inputs the first keyword.

  The weight acquisition unit 17 acquires the weight of the first keyword received by the keyword reception unit 13 from the weight storage unit 35.

  The level discriminating unit 19 inputs the keyword (first keyword) newly received by the keyword receiving unit 13 and the keyword receiving unit 13 from the same user in the past predetermined period (for example, one year or all periods). A weighted average value with the received keyword (second keyword) is obtained, and the obtained weighted average value is set as the current knowledge level of the user. Further, the knowledge level of the user stored in the user information storage unit 31 is updated using the set knowledge level. Further, the number of second keywords is updated by adding the number of first keywords to the number of second keywords stored in the user information storage unit 31 in synchronization with or asynchronously with the update process. To do.

  The answer output means responds to the knowledge level set by the level determination means 19 from the answer storage means by the cooperation of the answer acquisition means 21, the supplementary information acquisition means 23, and the answer determination means 25 as follows. Is output.

  The answer acquisition unit 21 asks a set of questions and answers related to the first keyword based on the question / answer search information including the first keyword received by the keyword receiving unit 13 from the question answer storage unit 41. The answer is retrieved and obtained from the answer storage means 41, and the obtained answer is stored in the answer temporary storage means 39. In the first embodiment, a question obtained by such a search is estimated as a question that the user wants to solve.

  The supplementary information acquisition unit 23 acquires supplementary information corresponding to the knowledge level set by the level determination unit 19 from the supplementary information storage unit 43 as information for supplementing the answer acquired by the answer acquisition unit 21.

  The answer determination means 25 updates the contents of the answer acquired by the answer acquisition means 21 (the answer stored in the answer temporary storage means 39) using the supplementary information acquired by the supplementary information acquisition means 23 from the supplementary information storage means 43. Then, it is determined as an answer corresponding to the knowledge level set by the level discriminating means 19 and output to the user terminal 3 so that it can be displayed.

  FIG. 10 is a sequence diagram showing the flow of answer processing in the user terminal 3 and the answering device 5. Hereinafter, the flow of the answer process will be described with reference to this figure.

  First, the information input / output unit 7 of the user terminal 3 transmits a UID, a predetermined password, and the like as information necessary for login (login information) to the answering device 5 based on a user instruction (step 101).

  After the log-in process is executed based on the received log-in information in the answering device 5, the level acquisition unit 15 obtains the knowledge level corresponding to the UID included in the log-in information from the user information storage unit 31 and the first level for the log-in user. The number of keywords (number of search terms) of 2 is acquired (step 102).

  In this example, it is assumed that a user having a UID “U01” has logged in, and the level acquisition unit 15 acquires a value of 5 as the knowledge level and 4 as the number of second keywords. .

  Further, when receiving the user login, the answering device 5 transmits screen information (such as an html file) for searching for questions and answers as shown in FIG. 11 to the user terminal 3 (step 103).

  Subsequently, the information input / output means 7 of the user terminal 3 transmits the question / answer search information including the first keyword to the answering device 5 based on the user's instruction (step 104).

  For example, the user searches on a search screen (see FIG. 11) displayed on a display connected to the user terminal 3 for a question / answer search including a first keyword that he / she knows about a situation to be solved or information to be known. The question / answer search information can be transmitted to the answering device 5 by inputting the information and clicking the search button. In the illustrated example, as the question / answer search information, the AND search column and the OR search column include the first keyword related to (for example, included in) the search target and the NOT search column. The first keyword related (for example, included) to the answer to be excluded from the search target, the category to which the first keyword belongs in the category column, and the answer to be searched in the update column are stored in the question answer storage unit 41. The stored (updated) time can be input.

  In the answering device 5, when the keyword accepting means 13 obtains the question / answer search information including the first keyword from the user terminal 3 (step 105), the answer obtaining means 21 stores the question answer storage based on the question / answer search information. An answer to be searched and a question corresponding to the answer are searched from the means 41 (FIG. 5), and the obtained answer is stored in the answer temporary storage means 39 (step 106).

  Specifically, a CID is acquired from the classification storage means 33 (FIG. 3) based on the classification included in the question / answer search information, and a set of questions and answers corresponding to the acquired CID is searched. Next, from the searched question and answer set, the keywords ("memory" and "caution") entered in the AND search field of the first keyword are both of these keywords and the first keyword. Of these, the keywords ("addition" and "addition") entered in the OR search field include either one of the keywords, and do not include the first keyword (flash) entered in the NOT search field. Among them, the ones whose responses are stored within the period entered in the update column (after April 1, 2009) are extracted. A keyword or the like may be extracted from a sentence input as a free sentence and used.

  As a result, in the example shown in Fig. 5, the question "What do you need to be careful about when adding memory to a PC?" And "There are cases where all installed memory at the time of shipment is removed and replaced with purchased additional memory." Please remove the power cable, AC adapter, and battery before adding. "

  Next, the weight acquisition unit 17 acquires a weight from the weight storage unit 35 (FIG. 4) for each first keyword that has received an input (step 107).

  Specifically, the weight storage means 35 is referred to for each first keyword, and the weight “2, 0, 3, 0, 4” is assigned to the first keyword “memory, attention, expansion, addition, flash”. Can be obtained.

  When the weight acquisition unit 17 acquires the weight for each first keyword, the level determination unit 19 obtains the current knowledge level of the user who has input the question / answer search information including the first keyword, and the user information storage unit 31. The information on the user stored in (FIG. 2) is updated (step 108).

  Specifically, the level discriminating means 19 calculates a weighted average value indicating the current knowledge level of the user using the following equation.

  In the above formula, the old weighted average value is the weighted average value calculated for the user last time, that is, the knowledge level acquired from the user information storage means 31 in step 102, and the number of search terms (cumulative) is determined in the past. The number of keywords received from the user during the period, that is, the number of second keywords acquired from the user information storage means 31 in step 102, and the number of search terms (new) is the question received in step 105 / The number of first keywords included in the answer search information, and the Σ weight is the total weight for each keyword (first keyword) newly acquired in step 107.

  In this example, the old knowledge level is 5, the number of search terms (cumulative) is 4, the number of search terms (new) is 5, “memory, attention, expansion, addition, flash”, and the Σ weight is “3”. , 0, 4, 0, 5 "is 12. By substituting these values into the above equation, the current knowledge level becomes (5 × 4 + 12) / (4 + 5) ≈3.555 (4 by rounding), and as shown in FIG. The stored knowledge level is updated from 5 to 4, and the number of second keywords (number of search terms) is updated from 4 to 9.

  As described above, in the first embodiment, the current knowledge level of the user includes the keyword (second keyword) received from the user within the past predetermined period and the keyword (second keyword) newly received from the user. 1) and is set using the obtained weighted average value (the weighted average value is set as the current knowledge level). Thereby, if the weight of the first keyword newly input by the user is higher than the past knowledge level (for example, the latest knowledge level stored in the user information storage unit 31), If it is set higher and the weight of the first keyword newly input by the user is lower than the past knowledge level, it is set lower. Therefore, in the case of inputting a keyword that the user already knows as in the present embodiment, unlike the technique described in Japanese Patent Laid-Open No. 2008-21101, the first keyword is higher than the past knowledge level of the user. For example, since the current knowledge level of the user is assumed to be higher than the past knowledge level, the current knowledge level of the user can be set more accurately.

  Next, the supplementary information acquisition unit 23 acquires term supplementary information corresponding to the knowledge level set in step 108 from the supplementary information storage unit 43 (step 109).

  Specifically, first, the supplementary information acquisition unit 23 refers to the level reference storage unit 37 (FIG. 7), and acquires a level point corresponding to the current knowledge level of the user (4 set in step 108). In this example, if the knowledge level is 4, the level point is 1.

  Next, the supplementary information acquisition means 23 refers to the term field of the supplementary information storage means 43 (FIG. 6), and among the terms stored in the term field, answers to the answers obtained by the search in step 106. Identify the terms that are included. In this example, the two terms “installed memory” and “additional memory” among the terms stored in the term column are “removed and purchased all the installed memory at the time of shipment obtained in the search of step 106. It may be replaced with additional memory. Please remove the power cable, AC adapter, and battery before adding the memory card. ", Both of these terms are specified.

  Then, for each identified term, term supplement information corresponding to the obtained level point is obtained. In this example, since the level point is 1, "Installed memory (part of the memory can be assigned to the VRAM in the BIOS settings)" and "Additional memory (if there are two identical memories, dual channel can be enabled. Get additional terminology information.

  The answer determination means 25 updates the contents of the answer acquired by the answer acquisition means 21 in step 106 using the term supplement information acquired by the supplement information acquisition means 23 (step 110).

  Specifically, the answer determination means 25 obtains the “installed memory (a part of the memory can be allocated to VRAM by BIOS setting)” and “additional memory (the same memory is 2) acquired by the supplementary information acquisition means 23. Replacing the term supplementary information with the term "supplement target" in the answer stored in the answer temporary storage means 39 (Fig. 8) will update the contents of the answer. To do. In this example, the answer is updated from the content shown in FIG. 8 to the content shown in FIG. 9 by such replacement.

  Further, the answer determination means 25 determines the updated answer as an answer corresponding to the current knowledge level of the user set in step 108, and outputs the answer to the user terminal 3 together with the question searched in step 106. (Step 110).

  The information input / output means 7 of the user terminal 3 displays the question and answer received from the answering device 5 on the display (step 112).

  Specifically, as shown in FIG. 13, a question and answer in which term supplementary information is embedded in the answering device 5 is displayed. Since the answer corresponds to the current knowledge level of the user set in step 108, for example, as shown in FIG. 12, the question answer storage means 41 (see FIG. 12) regardless of the user's knowledge level. Compared with the case where the questions and answers stored in 5) are displayed as they are, a set of questions and answers can be presented more accurately.

The first embodiment of the present invention has been described above, but the present invention is not limited to the above-described embodiment, and can be variously modified and applied.
[Modification]
Unlike the first embodiment, the answering device 5 may be configured to accept input of question / answer search information from the user terminal 3 by e-mail, FAX, or telephone (eg, using IVR).

  When accepting input of question / answer search information by e-mail, for example, as shown in FIG. 15, a classification is specified from a character string (in this example, “computer”) input in the subject field, and the specified classification It is conceivable that a term (first keyword) included in the mail text is obtained by matching with reference to a predetermined attention term list. Here, in the attention term list, terms such as “computer”, “memory”, “hard disk”, “PC”, “CPU”, “addition”, “caution” are associated with the classification “computer”. If so, the first keywords “PC”, “Memory”, “Addition”, and “Caution” can be acquired from the illustrated mail text. Subsequent processing is the same as in the first embodiment, and it is conceivable to output an answer or a set of questions and answers to the user terminal 3 by a method of replying to an e-mail received from the user.

  When using FAX, by using a FAX receiver having an OCR function, question / answer search information can be specified from a character string printed on the received FAX and processed in the same manner as in the case of the above e-mail. . Also, when using FAX, by printing a FAX number indicating an answer output destination or an e-mail address, the OCR function is used to acquire the output destination information, and the user terminal (in this example, the user's FAX machine). Or a response to a personal computer).

  Further, when using a telephone, for example, an automatic voice response device (IVR [Interactive Voice Response]) is provided with the function of the answering device of the present invention and the function of automatic voice recognition. / Answer search information can be acquired and an answer can be output to a user terminal (in this example, a user's telephone) using an automatic response function. If different classifications are assigned to the numbers 0 to 9, it is possible to accept inputs based on the numbers pushed or dialed by the user.

  DESCRIPTION OF SYMBOLS 1 ... Reply system, 3 ... User terminal, 5 ... Answer apparatus, 7 ... Information input / output means, 9 ... Processing part, 11 ... Memory | storage part, 13 ... Keyword reception means, 15 ... Level acquisition means, 17 ... Weight acquisition means, DESCRIPTION OF SYMBOLS 19 ... Level discrimination means, 21 ... Answer acquisition means, 23 ... Supplementary information acquisition means, 25 ... Answer determination means, 31 ... User information storage means, 33 ... Classification storage means, 35 ... Weight storage means, 37 ... Level reference storage means 39 ... Temporary answer storage means, 41 ... Question answer storage means, 43 ... Supplementary information storage means

Claims (6)

Means for receiving input of a first keyword related to a question from a user;
A weighted average value of the first keyword that has received the input and the second keyword that has received an input from the user in the past predetermined period is obtained, and the obtained weighted average value is used as the current knowledge level of the user. Means for setting;
An answering apparatus comprising: means for determining and outputting an answer corresponding to the set knowledge level from means for storing a plurality of types of answers according to the knowledge level for each question. The means for storing the answer is:
Means for storing a set of multiple questions and answers;
Means for storing information for supplementing the answer in correspondence with the knowledge level,
The means for outputting is
Means for acquiring answers to questions related to the first keyword that received the input from means for storing the plurality of questions and answer sets;
Means for acquiring supplementary information corresponding to the set knowledge level from means for storing information for supplementing the answer;
The response device according to claim 1, further comprising: a unit that updates the content of the acquired response using the acquired supplemental information and determines the response as a response corresponding to the set knowledge level. The means for setting the knowledge level calculates the weighted average value using the following equation:

In the above formula, the old weighted average value is the weighted average value calculated for the user last time, the number of search terms (cumulative) is the number of second keywords that have received the input, and the number of search terms (new) is the input The number of first keywords that received Σ, the Σ weight is the sum of the weights of the first keyword,
The answering device according to claim 1 or 2. An answer system including a user terminal and an answer device configured to be able to communicate with the user terminal,
The user terminal includes means for transmitting a first keyword related to a question input by a user to the answering device,
The answering device is:
Means for receiving input of the first keyword from the user terminal;
A weighted average value of the first keyword that has received the input and the second keyword that has received an input from the user in the past predetermined period is obtained, and the obtained weighted average value is used as the current knowledge level of the user. Means for setting;
An answer system comprising: means for determining and outputting answers corresponding to the set knowledge level from means for storing a plurality of types of answers according to the knowledge level for each question. Receiving an input of a first keyword related to a question from a user;
A weighted average value of the first keyword that has received the input and the second keyword that has received an input from the user in the past predetermined period is obtained, and the obtained weighted average value is used as the current knowledge level of the user. Steps to set,
And a step of determining and outputting an answer corresponding to the set knowledge level from means for storing a plurality of types of answers according to the knowledge level for each question.   A computer program for causing an information processing apparatus to execute the answering method according to claim 5.

Images