JP2008090117A - Learning system, similar question retrieval method, and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To present a similar question or text by assuring matching between comments of questions or between the comment and the text in order to retrieve and display the similar question and text from the questions. <P>SOLUTION: A system accesses question information stored in a learning system server 20 from a user terminal 10 that is connectable to a network such as the Internet, and transmits instructions for similar question retrieval. The learning system server 20 calculates similarity between question and comment information in the stored question information to thereby extract the similar question, and transmits the similar question to the user terminal 10. According to this means, it is not required to manually and previously create indexes associated with questions to be retrieved, and the similar question can be retrieved without entering keywords. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a learning system, and more particularly to a learning system capable of searching for similar problems by similarity between commentary sentences.

  Conventionally, there has been a system for searching for a problem from a keyword. However, it is necessary to set a keyword for a problem to be searched for in advance, and a user also has to search by inputting a keyword.

  An example of a conventional learning system is described in Japanese Patent Laid-Open No. 2000-284679 (Patent Document 1). This conventional learning system is a system that analyzes a question sentence and presents a question sentence similar to the question sentence from the accumulated question answer database.

  However, there are cases where the route to the optimal solution is different depending only on the similarity of the question and the question sentence, and there is a limit in comparing the degree of similarity particularly from the test problem such as a mock test.

  In the answer commentary, information that is full of know-how such as solution techniques such as teachers and lecturers and prep schools and cram schools is accumulated. This system compares similarities of commentary information, searches for similar problems from the viewpoint of similar solutions, and further improves the learning effect with similar problems from learning one solution, and sets keywords for search in advance. There was no unnecessary learning system.

  Further, as a related technique, Japanese Patent Laid-Open No. 2002-366554 (Patent Document 2) discloses a search system for specifying a page such as a book, magazine, or information magazine using a network-connectable terminal such as a mobile phone. ing. This search system accesses a database from a terminal that can be connected to a network such as the Internet, and outputs data specifying a corresponding page such as a book, a magazine, or an information magazine at hand of an end user.

Japanese Patent Laid-Open No. 2001-357068 discloses an internet-based method that facilitates networking between interested persons and facilitates collaborative search for information.
The prior art is a method in a computer network for identifying users with similar interests, accepting a first query from a first user and storing a first item of information related to the first query. Using a second item of information, receiving a second query from a second user, storing a second item of information associated with the second query, and using the first item of information and the second item of information Calculating a similarity measure for the first query and the second query, comparing the similarity measure with a threshold similarity criterion, and determining whether the similarity measure is a threshold similarity determination. Informing the first user about the second user if criteria are met.

Japanese Patent Laid-Open No. 2003-066823 (Patent Document 4) discloses a communication correction system, a server, a terminal device, and a recording medium.
In this conventional technology, the student's information terminal activates a test question from the student question information database, displays a test start screen and a test screen, solves the problem, performs scoring, and displays a scoring screen as a scoring result. indicate. The information terminal transmits the scoring result information to the server, and receives similar question information of the wrong answer and answer explanation information corresponding to the scoring result information from the server. The information terminal displays comment information of the corresponding problem on the answer explanation (correct answer) screen, answer explanation (wrong answer) screen, and the like by the “explain” button. On the answer explanation (incorrect answer) screen, a “solve similar problem” button is displayed. By pressing “solve similar problem”, the information terminal receives the similar problem information received from the server on the weak point reinforcement problem screen. , Provide similar questions for wrong answers, and score similar questions. In addition, the information terminal displays an evaluation chart of test results on the evaluation screen by an “evaluation” button.

Japanese Patent Laid-Open No. 05-188849 (Patent Document 5) discloses an electronic learning machine.
In this electronic learning machine, the reference problem, its similar problems, and their answers are stored in the first storage device, and the learner outputs the reference problem and answers it, for example. The correctness / incorrectness of the answer is stored in the second storage device, and at the same time, a question pointer corresponding to the similar question to be next questioned is set. When the learner newly gives an instruction to start learning, the control device searches the second storage device, selects a similar question from the first storage device according to the set question pointer of each question, and notifies the learner. Present.

JP 2000-284679 A JP 2002-366554 A JP 2001-357068 A JP 2003-066683 A Japanese Patent Laid-Open No. 05-188849

An object of the present invention is to provide a learning system in which a learner can search for a similar problem based on the degree of similarity between answer commentary sentences without having to input a keyword.
Another object of the present invention is to provide a learning system for indicating similar problems and texts by taking a matching between the explanation sentences of the problems or between the explanation sentences and the text in order to search and display similar problems and texts from the problems. That is.

  In the following, means for solving the problem will be described using the numbers used in [Best Mode for Carrying Out the Invention] in parentheses. These numbers are added to clarify the correspondence between the description of [Claims] and [Best Mode for Carrying Out the Invention]. However, these numbers should not be used to interpret the technical scope of the invention described in [Claims].

The learning system of the present invention includes a user terminal (10) that displays options for searching for a question sentence, and a server (20) that stores the question sentence and the answer explanation sentence in association with each other.
The server (20) provides a question sentence and an answer explanation sentence to the user terminal (10), and another answer explanation sentence similar to the answer explanation sentence in response to a request for a similar question from the user terminal (10). And a question sentence corresponding to another answer explanation sentence is provided to the user terminal (10) as a similar problem.

  The user terminal (10) displays “type”, “year”, “exam”, and “subject” as attribute information of the question sentence as options.

  The user terminal (10) displays a question sentence in response to an operation on the option, displays an answer commentary in response to an instruction to display an answer, and receives a similar question received from the server (20) in response to a request for a similar question. indicate.

  The server (20) determines the similarity between the answer explanation sentences based on a predetermined function or a predetermined syntax included in the answer explanation sentences.

  The server (20) calculates and holds the similarity between the answer commentary sentences in advance.

  The server (20) extracts answer explanations of problems that could not be answered correctly based on past examination information and grade information, and searches for similar problems that could not be answered correctly.

  The similar problem search method of the present invention includes: (a1) a step of accessing a predetermined homepage from the user terminal (10); and (a2) a server (20) in response to an access to the predetermined homepage from the user terminal (10). ) Transmits the question information to the user terminal (10), and (a3) the user terminal (10) displays various question information on the screen, and when the source information is selected from the screen, A step of displaying detailed question information of the answer commentary; and (a4) when a “similar question search” button is operated, search information including the source information of the referenced question is transmitted to the server (20). (A5) When the server (20) receives the search information, the step (a5) extracts the answer explanation sentence based on the stored source information and question information; and (a6) the server (20) A step of searching for similarities with other stored answer explanations based on the answer explanations that have been issued; and (a7) the server (20) is configured to obtain high similarity question information obtained as a result of the search. , Transmitting to the user terminal (10), and (a8) the user terminal (10) displaying question information with a high degree of similarity.

  The similar problem search method of the present invention further includes a step (b1) in which the server (20) calculates and holds the similarity between the answer explanation sentences in advance.

  In the similar problem search method of the present invention, (c1) the server (20) extracts answer explanations of problems that could not be correctly answered based on the examination information and results information of past examinations, The method further includes searching for similar problems.

  The similar problem search method according to the present invention further includes a step (d1) in which the server (20) determines the similarity between the answer comment sentences based on a predetermined function or a predetermined syntax included in the answer comment sentence. It has.

  The learner can search similar problems with a click of a button with respect to the problem being referred to without inputting a keyword. Alternatively, similar problems can be searched from a problem number with a single button click. Further, only the question sentence, the answer explanation sentence, and the source information are stored in the learning system server, and it is not necessary to set a keyword for searching in advance. Furthermore, since similarities between answer commentary sentences are compared, similar problems can be retrieved and extracted from the viewpoint of similar solutions.

  The learning system of the present invention is a learning system that shows similar problems and texts by taking matching between the explanation sentences of the problems or between the explanation sentences and the text in order to search and display similar problems and texts from the problems.

A first embodiment of the present invention will be described below with reference to the accompanying drawings.
Referring to FIG. 1, the learning system of the present invention includes a user terminal 10 and a learning system server 20. The user terminal 10 and the learning system server 20 operate by program control and are connected to each other via a network 100 such as the Internet.

  The user terminal 10 is an information processing apparatus such as a personal computer, a mobile phone, a home game machine, or an Internet TV, and accesses the question information provided by the learning system server 20 and displays the problem and explanation on the screen. It has a function to do. The question information includes information on examination questions and explanations of answers to examination questions, and includes source information such as question numbers and pages, question school names, trial names, and trial rounds. In addition, the user terminal 10 has a function of transmitting search information, which is information for a learner to search for a similar problem, to the learning system server 20 via the network 100 by a learner's operation.

  The learning system server 20 is an information processing apparatus such as a workstation server that provides a problem, and receives search information including source information such as a question number and the number of pages transmitted by the user terminal 10 via the network 100. After calculating the similarity between the stored answer explanation sentences based on the received search information, it has a function of extracting question information and source information with high similarity and transmitting them to the user terminal 10. At this time, the question sentence and the answer commentary are stored in association with each other so that the correspondence between them can be understood.

  Note that the learning system server 20 performs matching between the commentary of the problem or between the commentary and the text when calculating the similarity. Then, the number of common words and symbols included in the commentary is determined, and the similarity is calculated. The calculation of the similarity is automatically performed by character search or document comparison.

  Further, the similarity may be obtained based on whether a mathematical expression or grammar similar to the commentary is used. For example, by assigning a common classification number to similar mathematical expressions and grammars, it is possible to determine that commentary sentences containing mathematical expressions and grammars having the same classification number are similar. Furthermore, it is also possible to make a difference in the degree of similarity depending on how many mathematical formulas and grammars having the same classification number are included between the explanatory sentences.

  Here, an example of means for automatically determining whether mathematical expressions and grammar are similar will be described. In mathematical formulas, answer explanations containing the same function or formula are determined to be similar. Further, in language grammar, it is determined that a predetermined idiom, wording, and answer explanation sentence including similar syntax are similar. At this time, characters and numbers (variables) that are not unique to mathematical formulas or grammars are matched using wild cards or regular expressions.

FIG. 2 shows a configuration example of a screen displayed on the user terminal 10.
This screen also serves as a search screen. Based on the search information designated by the user, the problem and commentary are displayed in the corresponding column. Here, the left side of the screen is a search information input field, and the right side of the screen is a search result display field.

  On the left side of the screen, items such as “type”, “year”, “examination”, “subject”, and “detail” are displayed based on the question information. “Type” and “Details” are displayed so that they can be selected by radio buttons or check boxes. The “year”, “examination”, and “subject” are displayed so that they can be selected by a selection part (select box). However, these are merely examples, and in actuality, it is only necessary that each of “type”, “year”, “examination”, “subject”, and “detail” is displayed so that it can be selected from options.

  “Type” indicates the type of test. Here, “Mock Test”, “University Entrance Exam Center Test”, and “University Entrance Test” are shown as options, and “Mock Test” is selected. “Year” indicates the year in which the test was conducted, and “2004” is selected here. “Test” indicates the specific content of the test. In this example, “first mark imitation test” is selected. “Subject” indicates a subject of the examination, and “English 1” is selected here. “Details” indicates details of the contents to be displayed. Here, “all questions”, “question number designation”, and “page designation” are shown as options, and “all questions” is selected. If “question number designation” or “page designation” is selected, the user can directly designate the question number or page. At this time, a question number that can be specified and an effective range of the page may be displayed as a guide when inputting. As another means, it is also possible to provide options for question numbers and pages and specify them by selection.

  On the right side of the screen, based on the search information, "2004 1st Mark Test English 1" is shown, and in the "Problem / answer comment display area", information such as test questions and test question answer sentences Is displayed. Furthermore, a link display and a button such as “display of answer / explanation” and “search similar problem” are displayed. For example, in the “Problem / answer comment display area”, only the test questions are initially displayed. When the user performs an operation on the link display labeled “Display of answers / comments”, It is possible to display the answer explanation text of the test question in the “sentence / answer explanation display area”. Also, by pressing a button labeled “Search similar questions”, it is possible to display a question similar to the test question in the “Problem / answer comment display area”.

  That is, as shown in FIG. 3, in the learning system of the present invention, “question sentence” is displayed on the user terminal 10 based on the search information, and then “indication of answer / explanation” is displayed according to the instruction “ "Answer commentary" is displayed. Here, when the learning system server 20 receives an instruction of “search for similar questions”, the learning system server 20 searches for “similar answer explanations” similar to the currently displayed “answer explanations”. The “similar question sentence” that is the question sentence of the “explanatory sentence” is presented. In addition, when there are a plurality of items that correspond to “similar question sentences”, all of them may be displayed, or from among the corresponding question sentences, one that is particularly desired to be displayed may be selected.

  Note that the “question sentence” and the “answer commentary sentence” are separate. The present invention is characterized in that when a user searches for a similar “question sentence”, the similar “question sentence” is presented based on the similarity of the “answer commentary sentence” prepared in advance.

  FIG. 4 shows an example of a data table of question information. The data table of question information stores “type”, “year”, “examination”, “subject”, “question number”, “page”, “question sentence”, and “answer commentary sentence”. However, in practice, it is not necessary that all items are stored in one data table. Each item is stored in an individual data table and linked as an RDB (relational database) based on key data. Anyway. Regarding the similarity between “answer commentary sentences”, for example, the similarity is determined by comparing the contents included in the “answer commentary sentence”. If necessary, the “classification number” between the answer explanation sentences is stored in the question information data table, and it is determined that the “answer explanation sentences” having the same “classification number” are similar to each other. May be.

  In comparison between “problem sentences”, even if a common word or sentence is included, it is not always a similar problem. However, in the comparison between the “answer commentary texts”, if the common words and sentences are included, there is a high possibility that the same theme and field are handled. Since the question sentences in which the “answer commentary texts” are similar to each other are “problem sentences” that are truly highly similar in learning, in the learning system, the “answer commentary” is more than the similarity of the “question sentence”. The similarity of “sentence” is more important. When you answer a specific question by mistake, you can learn that you can use the same answer and solution even if the question sentence changes by touching many answers and solutions with similar contents.

With reference to FIG. 5, the example of an internal structure of the user terminal 10 and the learning system server 20 is demonstrated.
The user terminal 10 includes a processing unit 11, a communication unit 12, a storage unit 13, and a display unit 14. The processing unit 11 controls each unit in the user terminal 10 and processes various information. The communication unit 12 accesses question information provided by the learning system server 20 via the network 100. The storage unit 13 stores information processed by the processing unit 11 and question information acquired from the learning system server 20. The display unit 14 displays information on the test questions and answer explanations of the test questions based on the question information stored in the storage unit 13.

  The learning system server 20 includes a processing unit 21, a communication unit 22, and a database 23. The processing unit 21 controls each unit in the learning system server 20 and processes various information. The communication unit 22 receives the search information transmitted from the user terminal 10 via the network 100. The database 23 stores information processed by the processing unit 21, question information, and search information received from the user terminal 10. The database 23 may be a storage server provided outside the learning system server 20.

Next, the operation of the first embodiment of the present invention will be described in detail with reference to the flowchart of FIG.
(1) Step S101
First, the learner accesses the homepage established by the learning system provider from the user terminal 10. Here, this homepage is provided by the learning system server 20. However, in practice, this home page can be provided by another server that can access the learning system server 20.
(2) Step S102
In response to access to the home page from the user terminal 10, the learning system server 20 transmits question information to the user terminal 10. Here, the question information is transmitted from the communication unit 22 to the communication unit 12 via the network 100.
(3) Step S103
On the user terminal 10, various question information is displayed on the screen, and the learner specifies unique source information such as the year of question, the university at which the question is given, and the trial round, and so on. Refer to detailed question information. These operations are performed by the processing unit 11.
(4) Step S104
Further, if a similar problem is to be searched, the “search similar problem” button is clicked with the mouse, and the key information of the referenced question is transmitted to the learning system server 20 as search information. Here, the search information is the content selected by the user on the above-described screen, and is transmitted from the communication unit 12 to the communication unit 22 via the network 100.
(5) Step S105
The learning system server 20 receives search information including source information from the user terminal 10.
(6) Step S106
When the learning system server 20 receives the search information, the learning system server 20 extracts an answer commentary based on the stored source information and question information. Here, the source information, the question information, and the answer commentary are associated and stored in the database 23.
(7) Step S107
The learning system server 20 searches for similarities with other stored answer explanation sentences based on the extracted answer explanation sentences. These operations are performed by the processing unit 21.
(8) Step S108
The learning system server 20 transmits question information with high similarity obtained as a result of the search and source information associated therewith to the user terminal 10. Here, the question information is transmitted from the communication unit 22 to the communication unit 12 via the network 100.
(9) Step S109
The user terminal 10 displays question information and source information for a problem similar to the problem specified when searching. The learner refers to the question sentence and the answer explanation sentence of the similar question by designating the source information displayed as the similar question. These operations are performed by the processing unit 11.

The second embodiment of the present invention will be described below.
The second embodiment of the present invention is different from the first embodiment described above in that the similarity between answer comment sentences stored on the learning system server is calculated and held in advance.

Next, the operation of the second exemplary embodiment of the present invention will be described in detail with reference to FIG.
First, when a learner accesses the learning system server 20 from the user terminal 10 via the network 100, in response to this, the learning system server 20 transmits question information and source information to the user terminal 10. When the learner specifies the source information, the question information is displayed on the screen on the user terminal 10. When the learner clicks the “similar question search” button, unique search information including source information is transmitted to the learning system server 20. The learning system server 20 receives the search information and extracts answer commentary information. The steps so far are the same as those in the first embodiment.

  In the learning system server 20, the similarity between each answer commentary is calculated and held in advance, and source information with a high similarity is stored. For the source information in the search information received from the user terminal, the stored source information having a high degree of similarity is transmitted to the user terminal 10.

  In the second embodiment of the present invention, since the similarity is calculated in advance, the search speed of the similar problem can be increased.

The third embodiment of the present invention will be described below.
When managing personal information and grade information of students and students at a cram school, prep school, school, etc., a mechanism to extract answer explanations of questions that could not be answered correctly based on past examination information and grade information Combine and search for similar questions that could not be answered correctly.

  In the third embodiment of the present invention, similar problems can be extracted more in line with the needs of individual learners by combining with grade information.

  The present invention can be applied to the use of a learning support system in places such as a prep school, a cram school, a school, a qualification acquisition school, and a company in-house education. It can also be applied to applications such as a self-study system using the Internet and the display of titles of similar problem publications for promoting reference book sales.

  As described above, the learning system of the present invention accesses the question information stored in the learning system server 20 from the user terminal 10 that can be connected to a network such as the Internet, and transmits a similar problem search instruction. The learning system server 20 calculates the similarity between the answer explanation information in the stored question information, extracts the similar problem, and transmits the similar problem to the user terminal 10. Thereby, it is not necessary to manually create an index associated with the search target problem in advance, and a similar problem can be searched without inputting a keyword.

FIG. 1 is a block diagram of the learning system of the present invention. FIG. 2 is a diagram illustrating an example of a screen displayed on the user terminal. FIG. 3 is a conceptual diagram showing the features of the present invention. FIG. 4 is a diagram illustrating an example of a data table of question information. FIG. 5 is a block diagram illustrating a configuration example of the user terminal and the learning system server. FIG. 6 is a flowchart showing the operation of the present invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 ... User terminal 11 ... Processing unit 12 ... Communication unit 13 ... Storage unit 14 ... Display unit 20 ... Learning system server 21 ... Processing unit 22 ... Communication unit 23 ... Database 100 ... Network

Claims (11)

A user terminal that displays options for searching for a problem sentence;
A server for storing the question sentence and the answer explanation sentence in association with each other;
The server provides the question sentence and the answer explanation sentence to the user terminal, and searches for another answer explanation sentence similar to the answer explanation sentence in response to a request for a similar question from the user terminal. A learning system that provides a question sentence corresponding to the other answer explanation sentence as a similar problem to the user terminal. The learning system according to claim 1,
The user system displays “type”, “year”, “exam”, and “subject” as attribute information of the question sentence as the options. The learning system according to claim 1 or 2,
The user terminal displays the question sentence in response to an operation on the option, displays the answer commentary in response to an instruction to display an answer, and displays the similar problem received from the server in response to a request for a similar problem. Learning system to display. The learning system according to any one of claims 1 to 3,
The server determines a similarity between the answer comment sentences based on a predetermined function or a predetermined syntax included in the answer comment sentence. The learning system according to any one of claims 1 to 4,
The server is a learning system in which the similarity between the answer commentary is calculated and held in advance. The learning system according to any one of claims 1 to 5,
The said server extracts the answer commentary of the problem which was not able to answer correctly based on the examination information and results information of the past examination, and searches the similar problem of the problem which could not be answered correctly. (A1) accessing a predetermined homepage from the user terminal;
(A2) in response to access from the user terminal to the predetermined homepage, the server transmits question information to the user terminal;
(A3) The user terminal displays various question information on a screen, and when source information is selected from the screen, displays detailed question information of a question sentence and an answer commentary;
(A4) When the “search for similar problem” button is operated, the search information including the source information of the question being referred to is transmitted to the server;
(A5) When the server receives the search information, the server extracts the answer commentary based on the stored source information and question information;
(A6) the server searching for a similarity to another stored answer explanation sentence based on the extracted answer explanation sentence;
(A7) the server transmits the question information having a high similarity obtained as a result of the search to the user terminal;
(A8) The similar problem search method, wherein the user terminal includes a step of displaying the question information with the high similarity. The similar problem search method according to claim 7,
(B1) The similar problem search method further comprising a step in which the server calculates and holds the similarity between the answer comment sentences in advance. The similar problem search method according to claim 7 or 8,
(C1) The server further includes a step of extracting an answer explanation sentence of a problem that could not be correctly answered based on examination information and results information of past examinations, and searching for a similar problem of the problem that could not be correctly answered. Similar problem search method. The similar problem search method according to any one of claims 7 to 9,
(D1) The similar problem search method further comprising the step of the server determining a similarity between the answer commentary sentences based on a predetermined function or a predetermined syntax included in the answer commentary sentence.   A program for causing a computer to execute the similar problem search method according to any one of claims 7 to 10.

Images