Learning support device, learning support method, and program

Abstract

To provide an information processing device, an information processing method, and a program capable of easily giving questions of appropriate difficulty corresponding to a learning level of a user.SOLUTION: An information processing device is provided with a processing part for identifying an approximate user indicating that a certain object user among a plurality of users approximates a proficiency trend of a plurality of questions from among the plurality of users on the basis of the existence / nonexistence of a question giving history to the plurality of users and question giving history information including correctness determination results of given questions about the plurality of questions, deriving a prediction accuracy rate of the object user to prediction object questions being at least ones of ungiven questions that are not given to the object user and wrong answer questions wrongly answered by the object person in the past among the plurality of questions on the basis of the correctness determination results of the prediction object problems of the approximation user in question giving history information, and determining questions to be given to the object user among the plurality of problems on the basis of the derived prediction accuracy rate.SELECTED DRAWING: Figure 13

Description

Technical Field

[0001] The present invention relates to an information processing apparatus, an information processing method, and a program.

Background Art

[0002] Conventionally, there is a terminal device that can be used for learning various subjects such as languages and can conduct tests related to the subject (question posing and correct / incorrect judgment of answers). In such a test, by posing questions with a difficulty level corresponding to the user's learning level, it is possible to appropriately determine the user's learning level or enhance the learning effect of the test itself.

[0003]

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] However, since the above prior art assumes that the degree of understanding of a concept is reflected in the correct answer rates of other questions, it cannot be applied when this assumption does not hold (for example, in the case of a word test where the questions are not combinations of multiple concepts). Also, even if the above assumption holds, since the questions are not necessarily composed only of simple combinations of concepts, the prediction accuracy of the correct answer rate is not necessarily high. ​Thus, when using the above prior art, there are limitations in applying the method for predicting the correct answer rate, and the prediction accuracy of the correct answer rate is not necessarily high. Therefore, there is a problem that it is not easy to conduct an appropriate difficulty test according to the user's learning level.

[0006] An object of the present invention is to provide an information processing apparatus, an information processing method, and a program that can easily present problems of an appropriate difficulty according to the user's learning level.

Means for Solving the Problem

[0007] To solve the above problems, an information processing apparatus according to the present invention Based on the question history information including the presence or absence of the question history for a plurality of users regarding a plurality of questions and the correct / incorrect determination results of the questions that have been presented, a target user among the plurality of users and an approximate user whose familiarity tendency with the plurality of questions is approximated are specified from among the plurality of users. Among the plurality of questions, the predicted correct answer rate of the target user for a predicted target question that is at least one of an unpresented question that has not been presented to the target user and a wrong answer question that the target user has answered incorrectly in the past is derived based on the correct / incorrect determination result of the predicted target question of the approximate user in the question history information. It includes a processing unit that determines a question to be presented to the target user among the plurality of questions based on the derived predicted correct answer rate. It is characterized by this.

[0008] Also, to solve the above problems, an information processing method according to the present invention Is an information processing method executed by a computer of an information processing system, Based on the question history information including the presence or absence of the question history for a plurality of users regarding a plurality of questions and the correct / incorrect determination results of the questions that have been presented, a target user among the plurality of users and an approximate user whose familiarity tendency with the plurality of questions is approximated are specified from among the plurality of users. Of the plurality of problems, the predicted correct answer rate of the target user for a predicted target problem that is at least one of an unasked problem that has not been asked to the target user and a wrong answer problem that the target user has answered incorrectly in the past is derived based on the correct / incorrect determination result of the predicted target problem of the approximate user in the question history information, Based on the derived predicted correct answer rate, a problem to be asked to the target user is determined from among the plurality of problems, The determined problem is asked to the target user characterized by.

[0009] Also, in order to solve the above problems, the program according to the present invention is On a computer provided in an information processing apparatus, Based on question history information including the presence or absence of question history for a plurality of users regarding a plurality of questions and the correct / incorrect determination results of the questions that have been asked, a process of specifying an approximate user whose learning tendency regarding the target user and the plurality of questions is approximate from among the plurality of users, A process of deriving the predicted correct answer rate of the target user for a predicted target problem that is at least one of an unasked problem that has not been asked to the target user and a wrong answer problem that the target user has answered incorrectly in the past based on the correct / incorrect determination result of the predicted target problem of the approximate user in the question history information, A process of determining a problem to be asked to the target user from among the plurality of problems based on the derived predicted correct answer rate characterized by causing to execute.

Effect of the Invention

[0010] According to the present invention, it is possible to easily conduct a test with a difficulty level corresponding to the learning level of the user.

Brief Description of the Drawings

[0011]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

Figure 15

Figure 16

Figure 17

Figure 18

Figure 19

Mode for Carrying Out the Invention

[0012] Hereinafter, embodiments of the present invention will be described based on the drawings.

[0013] <Configuration of the Learning Support System> FIG. 1 is a schematic configuration diagram of the learning support system 1 of the present embodiment. The learning support system 1 (information processing system) includes a server 10 (information processing device) and a plurality of terminal devices 20 that are communicably connected to the server 10 via a communication network N. The communication network N is, for example, the Internet, but is not limited thereto, and may be another network such as a LAN (Local Area Network). At least a part of the communication path between the server 10 and the terminal device 20 may be a wireless communication path.

[0014] The learning support system 1 provides a learning support service for assisting the language learning of users who use the terminal device 20. The terminal device 20 is, for example, a smartphone, but is not limited thereto, and may be a tablet-type terminal, a notebook PC (personal computer), or a stationary PC.

[0015] A learning application program (hereinafter referred to as "learning app 231 (see FIG. 6)") is installed in the terminal device 20. By executing this learning app 231, the terminal device 20 provides various services related to language learning to the user in cooperation with the server 10. For example, when a search instruction for a word (dictionary headword, dictionary item) is input from the user during the execution of the learning app 231, the terminal device 20 acquires and displays item information including the meaning and example sentences of the word from the server 10. In addition, the terminal device 20 can execute a word test for measuring the proficiency of words during the execution of the learning app 231. The word test is, for example, a question for answering a translation for the spelling (notation) of a word, or a question for answering a spelling for the translation of a word. When executing the test, the terminal device 20 acquires a list of words to be presented, and data on the notation and translation of each word from the server 10, presents the questions to the user, and when an answer to the question is input from the user, determines the correctness of the answer and presents the result to the user. The above is an example of the services provided by the learning support system 1 and is not limited thereto.

[0016] The learning support system 1 provides learning support services to a plurality of users who each use a plurality of terminal devices 20. The server 10 manages information related to the usage status of the learning support services for each user, and provides an appropriate service to the user according to the information. For example, when analyzing the question-giving status of word tests to a plurality of users and conducting a word test for a certain user, based on the above analysis results, questions with an appropriate difficulty level corresponding to the user's learning level are given. The method of determining the questions to be given will be described in detail later.

[0017] <Configuration of the server> FIG. 2 is a block diagram showing the functional configuration of the server 10. The server 10 includes a CPU 11 (Central Processing Unit), a RAM 12 (Random Access Memory), a storage unit 13, an operation unit 14, a display unit 15, a communication unit 16, a bus 17, and the like. Each part of the server 10 is connected via the bus 17.

[0018] The CPU 11 is a processor (processing unit) that reads and executes the server control program 131 (program) stored in the storage unit 13 and performs various arithmetic processes to control the operation of the server 10. Note that the server 10 may have a plurality of processors (for example, a plurality of CPUs), and the plurality of processes executed by the CPU 11 in the present embodiment may be executed by the plurality of processors. In this case, the processing unit is constituted by the plurality of processors. In this case, the plurality of processors may be involved in common processing, or the plurality of processors may independently execute different processes in parallel.

[0019] The storage unit 13 is a non-temporary recording medium readable by the CPU 11 as a computer, and stores the server control program 131 and various data. The storage unit 13 includes a non-volatile memory such as an HDD (Hard Disk Drive) or an SSD (Solid State Drive). The server control program 131 is stored in the storage unit 13 in the form of program codes readable by a computer. The data stored in the storage unit 13 includes a user management DB (database) 132 (feature information), a dictionary DB 133, a learning history DB 134, a question history DB 135 (question history information), and a predicted correct answer rate DB 136, etc.

[0020] FIG. 3 is a diagram showing an example of the contents of the user management DB 132. The user management DB 132 stores data related to a plurality of users who use the learning support service. The user management DB 132 includes feature information related to at least one of the attributes and characteristics of each user. One data row (record) of the user management DB 132 corresponds to one user. In the example shown in FIG. 3, the user management DB 132 has data columns (columns) of "user ID", "grade", and "desired school". The "user ID" is a unique code assigned to each user. The "grade" is the grade of the user. When the user is not a pupil or a student, the data of "general" is registered. The "desired school" is the desired school pre-entered by the user. Note that the user management DB 132 may further include data columns related to user attributes (such as gender and age) and characteristics (such as study time, study time zone, classification of favorite words, etc.).

[0021] FIG. 4 is a diagram showing an example of the contents of the dictionary DB 133. One data row (record) of the dictionary DB 133 corresponds to one word (item) in the English-Japanese dictionary. The information included in one data row of the dictionary DB 133 corresponds to the "item information" related to one item. The dictionary DB 133 has data columns (columns) of "word ID", "word", and "translation". The "Word ID" is a unique code assigned to the word in that data row. The "word" is the notation (spelling) of the word. The "translation" is the Japanese translation of the word. Note that the dictionary DB133 may further store data related to dictionaries other than the English-Japanese dictionary (such as English-English dictionaries).

[0022] Figure 5 is a diagram showing an example of the contents of the learning history DB134. The learning history DB134 stores data related to the learning history of each user in the learning support service. The learning history DB134 has data blocks generated for each user. One data row (record) in each data block corresponds to the word searched by that user or the word presented to that user as a word test. Each data block has data columns (fields) of "User ID", "Word ID", "Search Date", "Question Date", and "Correct / Incorrect Judgment Result". The "User ID" is the same code as the user ID in the user management DB132 in Figure 3. The "Word ID" is the word ID of the word corresponding to that data row and is the same code as the word ID in the dictionary DB133 in Figure 4. The "Search Date" is the date representing the time when the word in that data row was last searched. Note that in addition to the date, time information may be further registered. The "Question Date" is the date representing the time when the word in that data row was last presented in the word test. Note that in addition to the date, time information may be further registered. The "Correct / Incorrect Judgment Result" is the judgment result of the correctness of the user's answer in the word test, where "1" represents a correct answer and "0" represents an incorrect answer. For example, in the learning history DB134 shown in Figure 5, it is registered that the user with the user ID "U00000" answered correctly in the word test on "August 12, 2021" for the word with the word ID "W0012" and then conducted a search on "August 15, 2021". Examples of the contents of the question history DB135 and the predicted correct answer rate DB136 will be described later.

[0023] The operation unit 14 shown in FIG. 2 has a pointing device such as a mouse and a keyboard, etc., accepts position input and key input by the user, etc., and outputs the operation information to the CPU 11.

[0024] The display unit 15 includes a display device such as a liquid crystal display, and performs various displays on the display device according to a display control signal from the CPU 11.

[0025] The communication unit 16 is composed of a network card or the like, and transmits and receives data to and from the terminal device 20 on the communication network N according to a predetermined communication standard.

[0026] <Configuration of Terminal Device 20> FIG. 6 is a block diagram showing the functional configuration of the terminal device 20. The terminal device 20 includes a CPU 21, a RAM 22, a storage unit 23, an operation unit 24, a display unit 25, an audio output unit 26, a communication unit 27, a bus 28, etc. Each unit of the terminal device 20 is connected via the bus 28.

[0027] The CPU 21 is a processor (processing unit) that reads and executes programs such as the learning application 231 stored in the storage unit 23, and controls the operation of the terminal device 20 by performing various arithmetic processes. Note that the terminal device 20 may have a plurality of processors (for example, a plurality of CPUs), and a plurality of processes executed by the CPU 21 in the present embodiment may be executed by the plurality of processors. In this case, the processing unit is constituted by the plurality of processors. In this case, the plurality of processors may be involved in common processing, or the plurality of processors may independently execute different processes in parallel.

[0028] The storage unit 23 is a non-temporary recording medium readable by the CPU 21 as a computer, and stores programs such as the learning application 231 and various data. The storage unit 23 includes a non-volatile memory such as a flash memory, for example. The program is stored in the storage unit 23 in the form of computer-readable program code.

[0029] The operation unit 24 has a touch panel provided overlaid on the display screen of the display unit 25, physical buttons, etc., and accepts touch operations on the touch panel by the user, pressing operations on the physical buttons, etc., and outputs the operation information to the CPU 21.

[0030] The display unit 25 includes a display device such as a liquid crystal display, and performs various displays on the display device according to a display control signal from the CPU 21.

[0031] The audio output unit 26 has a speaker, and outputs voices such as the generation of words and test questions according to an audio output control signal from the CPU 21. Further, the audio output unit 26 outputs an audio signal to an external audio output device (for example, earphones or headphones, etc.) connected by wire or wirelessly, and causes the audio output device to output audio.

[0032] The communication unit 27 is composed of a communication module including an antenna, etc., and transmits and receives data to and from the server 10 on the communication network N according to a predetermined communication standard.

[0033] <Operation of the learning support system> Next, the operation of the learning support system 1 will be described. The operating entities in the operations shown below are the CPU 11 of the server 10 and the CPU 21 of the terminal device 20, but hereinafter, for convenience, the server 10 and the terminal device 20 may be described as the operating entities.

[0034] A user who receives a learning support service by the learning support system 1 executes a learning app 231 on a terminal device 20 and logs in to the learning support system 1. The login to the learning support system 1 is performed by an authentication process that collates, for example, a user ID and a password input by the user and transmitted from the terminal device 20 to the server 10 with the user ID and password registered in advance in the server 10. A user who has logged in to the learning support system 1 can execute various functions of the learning support service on the learning app 231.

[0035] (Word search) One of the basic functions of the learning support service by the learning support system 1 is the search for words in an English-Japanese dictionary. The user can cause a dictionary screen 40 for performing a word search in the English-Japanese dictionary to be displayed on the display unit 25 by performing a predetermined operation on the learning app 231.

[0036] FIG. 7 is a diagram showing the dictionary screen 40. On the dictionary screen 40, there are displayed a search box 41 for specifying a word to be searched, an item information display area 42 where item information of the searched word is displayed, a voice playback button 43 for playing back the voice of the word, a dictionary screen end button 44 for ending the word search, and the like. When a search is executed with the spelling of the word to be searched entered in the search box 41, the terminal device 20 requests the server 10 for the item information of the word. The server 10 acquires the item information of the specified word from the dictionary DB 133 in response to the request and transmits it to the terminal device 20. The terminal device 20 displays the acquired item information in the item information display area 42. Further, the server 10 registers the search date on which the word was searched in the data block corresponding to the user who performed the search in the learning history DB 134. Thereby, the word search history by the user is recorded.

[0037] (Test) In the learning support service provided by the learning support system 1, the above-described word test can be conducted. Hereinafter, a word test in which the user answers a translation for a word's spelling (notation) will be described as an example. In the word test of the present embodiment, among the words included as items in the dictionary DB 133, a word that the user has not searched for in the past and has never been presented in a past word test (an unpresented question) is selected as the presented word (hereinafter referred to as the "presented word"). When the user gives an instruction to execute a word test on the learning app 231, first, a difficulty level designation screen 50 for designating the difficulty level (level) of the word test is displayed on the display unit 25.

[0038] FIG. 8 is a diagram showing the difficulty level designation screen 50. On the difficulty level designation screen 50, a difficulty level designation button 51 for designating the difficulty level of the word test, a test start button 52 for starting the word test, and the like are displayed. In the present embodiment, any one of five levels of difficulty from the lowest "Level 1" to the highest "Level 5" can be designated by the difficulty level designation button 51.

[0039] When an operation is performed to select the test start button 52 in a state where any difficulty level is specified by the difficulty level specifying button 51, the terminal device 20 requests the server 10 for a list of question words (hereinafter referred to as the "question word list") that will be at the specified difficulty level. In response to the request, the server 10 obtains question words from the dictionary DB 133 that are at the difficulty level specified for the user and generates a question word list. In the prediction correct answer rate DB 136 in the storage unit 13 of the server 10, information on the prediction correct answer rate by the user is registered for words that have never been used as question words in the past word tests for the user. The server 10 extracts question words at the specified difficulty level based on this prediction correct answer rate. Specifically, as shown in FIG. 9, when the specified difficulty level is "level 5", question words are selected from among the words with a prediction correct answer rate of 0.0 or more and less than 0.2; when the specified difficulty level is "level 4", question words are selected from among the words with a prediction correct answer rate of 0.2 or more and less than 0.4; when the specified difficulty level is "level 3", question words are selected from among the words with a prediction correct answer rate of 0.4 or more and less than 0.6; when the specified difficulty level is "level 2", question words are selected from among the words with a prediction correct answer rate of 0.6 or more and less than 0.8; and when the specified difficulty level is "level 1", question words are selected from among the words with a prediction correct answer rate of 0.8 or more and 1.0 or less.

[0040] The number of words to be included in the question word list, that is, the number of words to be used as question words in one word test, is predetermined by setting and is 10 in this embodiment. The server 10 selects 10 words with a prediction correct answer rate corresponding to the specified difficulty level, creates a question word list, and transmits it to the terminal device 20. When the terminal device 20 obtains the question word list, it causes the display screen by the display unit 25 to transition to a test screen 60 for performing a word test.

[0041] FIG. 10 is a diagram showing the test screen 60. On the test screen 60, there are displayed a problem word 61, a text box 62 for inputting a translation as an answer, an answer button 63, etc. The problem word 61 is one word included in the application word list. The user can answer the question by performing an operation of inputting the translation of the problem word 61 into the text box 62 and selecting the answer button 63. When the answer for one problem word 61 is completed, the next problem word 61 included in the problem word list is displayed, and the user can continue to answer. When the answers for all (in this embodiment, 10) problem words 61 are completed, the correct / incorrect judgment results for all the problem words 61 are displayed on the display unit 25. Alternatively, each time an answer is given for one problem word 61, the correct / incorrect judgment result for that problem word 61 may be displayed.

[0042] Also, the server 10 records the words put in the problem word list as words that have been used in a test in the learning history DB 134. That is, in the data block corresponding to the user who has taken the word test in the learning history DB 134, a data row of the word put in the problem word list is added, and the test date is registered. Also, the correct / incorrect judgment results for each problem word are registered. Thereby, the history of word test questions for the user and the history of correct / incorrect judgment results are recorded.

[0043] (Method for calculating predicted correct rate) Next, the method for calculating the predicted correct rate described above will be explained. The server 10 executes a predicted correct rate calculation process for calculating the predicted correct rate of words that have not been used in a word test for all users (a plurality of users) at a predetermined frequency (for example, about once a month). In this embodiment, the words that have not been used in a test (unanswered questions) correspond to the "predicted target questions" that are the targets for calculating the predicted correct rate.

[0044] In the prediction accuracy calculation process, first, for all the words registered in the dictionary DB133 (in other words, all the questions (multiple questions) of the word test), data on the presence or absence of the question history for each user and the question history including the correct / incorrect determination results of the words that have been questioned are collected and registered in the question history DB135. The correct / incorrect determination results of the words that have been questioned are aggregated based on the correct / incorrect determination results in the tests conducted within the most recent predetermined period (for example, within the past three months). When a word has been questioned multiple times within the above-mentioned predetermined period, the correct / incorrect determination result at the last time of questioning is used.

[0045] FIG. 11 is a diagram showing an example of the content of the question history DB135. One data row (record) in the question history DB135 corresponds to one user. The question history DB135 has a number of data rows corresponding to all users (here, assume there are 30,000 users with user IDs: U00000 to U29999). Also, one data column (column) in the question history DB135 corresponds to one word. The question history DB135 has a number of data columns corresponding to all the words registered in the dictionary DB133 (here, assume there are 3,000 words with word IDs: W0000 to W2999).

[0046] The data corresponding to each word in each data row is either "0", "1", or "null (no data)". "0" indicates that the word has been questioned one or more times within the most recent predetermined period for the user corresponding to the data row, and the correct / incorrect determination result at the last time of questioning was "incorrect answer". "1" indicates that the word has been questioned one or more times within the most recent predetermined period for the user corresponding to the data row, and the correct / incorrect determination result at the last time of questioning was "correct answer". "null" indicates that the word has never been questioned for the user corresponding to the data row in the past. In this way, the question history DB135 includes information on the question status (presence or absence of the question history) for all users regarding all words and the correct / incorrect determination results.

[0047] Next, based on the question history in the question history DB 135 and the correct / incorrect judgment results, for each user, approximate users whose proficiency tendencies for words are similar among all other users are identified. Here, a collaborative filtering method is used.

[0048] Specifically, first, for one target user who is the target for identifying approximate users and each of all other users (hereinafter referred to as "comparison users"), feature vectors are respectively identified. The feature vector is a vector (here, a vector consisting of the said elements) that includes, as elements, data of words that have been questioned for both the target user and the comparison user among the data of all the words in the data row of the question history DB 135 (3000 pieces of data in FIG. 11), that is, words for which the values are "0" or "1" in the data rows of the target user and the comparison user.

[0049] For example, taking as an example the case where the total number of words is 10 and the data corresponding to the target user and a certain comparison user for these words is as follows ("-" represents "null"). Target user: 01--01110- Comparison user: 101--01-1- In this case, the data of the 1st, 2nd, 6th, 7th, and 9th words that have been questioned for both the target user and the comparison user are used as the elements of the feature vector. The feature vector a corresponding to the target user and the feature vector b corresponding to the comparison user are identified as follows. Feature vector a = (0, 1, 1, 1, 0) Feature vector b = (1, 0, 0, 1, 1)

[0050] Subsequently, the cosine similarity between the feature vectors of the target user and the comparison user is calculated by the following formula.

[0051]

Equation

[0052] The closer the cosine similarity is to "1", the more similar the characteristics of the target user and the comparison user (here, the word proficiency tendency) are. For a certain target user, this cosine similarity is calculated between the target user and all the remaining comparison users. The data group D1 in FIG. 11 represents the calculation results of the cosine similarity between the user with the user ID "U00000" as the target user and all other comparison users. Note that depending on the number of words for which at least one of the users is "null", the dimensionality of the feature vector may decrease, and it may not be possible to perform the comparison of proficiency tendencies using the cosine similarity with the desired accuracy. Therefore, users with the dimensionality of the feature vector less than a predetermined number may be excluded from the extraction target of approximate users.

[0053] Based on the calculated cosine similarity of the data group D1, approximate users whose word proficiency tendencies are similar to those of the target user are identified. For example, among a plurality of users, comparison users with a cosine similarity equal to or higher than a reference value (for example, 0.5 or higher) may be specified as approximate users of the target user. Alternatively, among a plurality of users, a predetermined reference number (for example, "top 10", etc.) of comparison users selected in descending order of cosine similarity may be specified as approximate users of the target user. Alternatively, among a plurality of users, a predetermined reference ratio (for example, "top 5%", etc.) of comparison users selected in descending order of cosine similarity may be specified as approximate users of the target user. In the example of FIG. 11, three comparison users selected in descending order of cosine similarity are regarded as approximate users, and as approximate users, three users with user IDs "U00001" to "U00003" (whose cosine similarity is shown in region A) are identified.

[0054] Next, for the words that have not been presented to the target user (the colored parts in Fig. 11, that is, the three words with word IDs "W0002" to "W0004"), based on the data of the correct / incorrect judgment status of the approximate users (region B in Fig. 11), the predicted correct rate of the target user (region C in Fig. 11) is calculated (derived). Specifically, for each word that has not been presented, the average correct rate of the specified multiple approximate users is calculated and used as the predicted correct rate of the target user. For example, for the word with word ID "W0002" in Fig. 11, since the correct / incorrect judgment results of all three approximate users are "1" (correct answer), the predicted correct rate is calculated as "1". Also, for the word with word ID "W0003", since one of the three approximate users is "1" (correct answer) and two are "0" (incorrect answer), the predicted correct rate is calculated as "0.33". Also, for the word with word ID "W0003", since one of the three approximate users is "1" (correct answer), one is "0" (incorrect answer), and one is "null" (not presented), the predicted correct rate is calculated as "0.50".

[0055] In this way, a data group D2 of the predicted correct rates for all words that have not been presented to the target user is generated. Note that in this embodiment, even for words that have not been presented to the target user, if the target user has searched for them, they will not be the target of presentation. Therefore, the predicted correct rate of the searched words is set to "null". Thus, in this embodiment, among the words that have not been presented, the words that have not been searched for correspond to the "predicted target questions". However, it is not limited to this, and it may be set that all words that have not been presented are the target of presentation regardless of the presence or absence of the search history. In this case, for words with a search history, even if they have not been presented, they are regarded as "predicted target questions", and the predicted correct rate is calculated and registered in the data group D2. The data group D1 and the data group D2 generated in the process of deriving the predicted correct rate may be included in the question history DB135, or may be stored in a storage area other than the question history DB135 in the storage unit 13.

[0056] Thereafter, for each of the remaining users as the target user, in the same manner as above, identification of approximate users and derivation of the prediction accuracy rate are performed. The data group D2 of the prediction accuracy rate for each user is registered in the prediction accuracy rate DB136.

[0057] FIG. 12 is a diagram showing an example of the content of the prediction accuracy rate DB136. One data row of the prediction accuracy rate DB136 corresponds to one user, and in each data column (column) corresponding to a plurality of words, the prediction accuracy rate (the value of the data group D2 in FIG. 11) for that word of the user is registered. Also, for the words for which questions have been asked and the words that have been searched for with respect to the user corresponding to the data row, it is set to "null". The prediction accuracy rate of the prediction accuracy rate DB136 is used for the process of selecting words with a prediction accuracy rate corresponding to the specified difficulty level as described above.

[0058] (Control Procedure of Prediction Accuracy Rate Calculation Process) Next, the control procedure of the prediction accuracy rate calculation process for calculating the prediction accuracy rate will be described. FIG. 13 is a flowchart showing the control procedure of the prediction accuracy rate calculation process. As described above, the prediction accuracy rate calculation process is executed at a predetermined frequency such as once a month.

[0059] When the prediction accuracy rate calculation process is started, the CPU 11 of the server 10 acquires the result data of the word tests of all users (step S101). Here, the CPU 11 acquires the words for which questions have been asked and the correct / incorrect determination results for those words from the data blocks of each user in the learning history DB134, and registers the content in the question history DB135.

[0060] The CPU 11 substitutes 0 for the variable N representing the ordinal number of the user (step S102). Hereinafter, the Nth user (in this embodiment, N ranges from 0 to 29999) will be referred to as "user N".

[0061] The CPU 11 determines whether the variable N is less than the total number of users (step S103). If it is determined that the variable N is less than the total number of users (i.e., "YES" in step S103), the CPU 11 calculates the cosine similarity between the user N (the target user) and each other user (the comparison user), and generates the data group D1 shown in FIG. 11 (step S104).

[0062] The CPU 11 extracts a predetermined number of users in descending order of the cosine similarity and identifies them as the approximate users of the user N (step S105). Note that the method for identifying approximate users based on the cosine similarity is not limited to this as described above.

[0063] The CPU 11 assigns 0 to the variable M representing the ordinal number of the word (step S106). Hereinafter, the M-th word (in this embodiment, M ranges from 0 to 2999) is denoted as "word M".

[0064] The CPU 11 determines whether the variable M is less than the total number of words (step S107). If it is determined that the variable M is less than the total number of words (i.e., "YES" in step S107), the CPU 11 determines whether the user N has searched for the word M (step S108). Here, the CPU 11 determines that the user N has searched for the word M when the search date is registered in the data row of the word M in the data block of the user N in the learning history DB 134.

[0065] If it is determined that the user N has not searched for the word M (i.e., "NO" in step S108), the CPU 11 determines whether the word M has been presented to the user (step S109). Here, the CPU 11 determines that the word M has been presented to the user N when the presentation date is registered in the data row of the word M in the data block of the user N in the learning history DB 134.

[0066] When it is determined that the word M has not been presented to the user N (i.e., "NO" in step S109), the CPU 11 calculates the average correct answer rate of approximate users for the word M as the predicted correct answer rate of the word M (step S110).

[0067] On the other hand, in step S108, when it is determined that the user has searched for the word M (i.e., "YES" in step S108), or in step S109, when it is determined that the word M has been presented to the user (i.e., "YES" in step S109), the CPU 11 sets the predicted correct answer rate for the word M to "null" (step S111). Note that if the searched words among the words not yet presented are also to be the subject of the presentation, the determination step in step S108 is omitted.

[0068] When step S110 or step S110 ends, the CPU 11 registers the calculation result of the predicted correct answer rate in the predicted correct answer rate DB 136 (step S112).

[0069] The CPU 11 increments the variable M (step S113) and returns the process to step S107. In step S107, when it is determined that the variable M has reached the total number of words (i.e., "NO" in step S107), the CPU 11 increments the variable N (step S114) and returns the process to step S103. In step S103, when it is determined that the variable N has reached the total number of users (i.e., "NO" in step S103), the CPU 11 ends the predicted correct answer rate calculation process.

[0070] (Control Procedure of Test Processing) Next, the control procedure of the test processing for performing the word test will be described. FIG. 14 is a flowchart showing the control procedure of the test processing. In FIG. 14, the test processing executed by the CPU 21 of the terminal device 20 and the test processing executed by the CPU 11 of the server 10 are shown together.

[0071] When the test process starts, the CPU 21 of the terminal device 20 causes the display unit 25 to display the difficulty level specification screen 50 (step S201).

[0072] The CPU 21 determines whether an operation for specifying the difficulty level (the operation of selecting the test start button 52 in a state where the difficulty level specification button 51 in FIG. 8 is selected) has been performed (step S202). If it is determined that the operation has not been performed (in step S202, “NO”), step S202 is executed again. If it is determined that the operation for specifying the difficulty level has been performed (in step S202, “YES”), the CPU 21 requests the server for the question word list of the specified difficulty level (step S203). Here, the CPU 21 transmits a request signal for the question word list to the server 10.

[0073] When the CPU 11 of the server 10 receives the request signal for the question word list, it refers to the data row corresponding to the user (target user) during the execution of the test in the prediction correct answer rate DB 136, and extracts the words of the prediction correct answer rate (see FIG. 9) corresponding to the specified difficulty level (step S301). Further, the CPU 11 generates data of a question word list including the extracted words by a predetermined number (10 in this embodiment) and including data of the notation and translation of each word, and transmits it to the terminal device 20 (step S302).

[0074] When the CPU 21 of the terminal device 20 receives the question word list, it causes the display unit 25 to display the test screen 60 and starts the word test (step S204). Here, the CPU 21 displays one of the words included in the question word list on the test screen 60, accepts the input of the user's translation answer in the text box 62, and when the answer button 63 is selected in a state where the translation has been input, performs a correct / incorrect determination by comparing with the translation data of the question word list. Thereafter, this process is executed for all the words included in the question word list. When answers from the user are made for all the words, the CPU 21 causes the display unit 25 to display the correct / incorrect determination results for all the words.

[0075] The CPU 21 determines whether the word test has ended (whether the above-described correct / incorrect determination result has been displayed) (step S205). If the CPU 21 determines that the word test has not ended (in step S205, “NO”), it executes step S205 again. If it determines that the word test has ended (in step S205, “YES”), it ends the test process.

[0076] On the other hand, the CPU 11 of the server 10 registers the question date in the learning history DB 134 for the words included in the question word list transmitted in step S302 (step S303). Thereby, the question history of the word for the user is recorded. Also, in the data row corresponding to the user during the execution of the test in the prediction correct answer rate DB 136, the data corresponding to the word that has been questioned is changed to “null” (step S304). Thereby, in the test after the next time, the words that have already been questioned are not extracted as question words. When step S304 ends, the CPU 11 ends the test process.

[0077] <Modification Example 1> Next, Modification Example 1 of the above embodiment will be described. In this modification example, the method for specifying approximate users is different from that of the above embodiment, and other points are the same as those of the above embodiment. Below, the differences from the above embodiment will be described.

[0078] In the above embodiment, the correct / incorrect determination results of the words that have been questioned in the question history DB 135 (question history information) are used as the feature vectors of the users used for calculating the cosine similarity. In contrast, in this modification example, approximate users are specified based on the correct / incorrect determination results of the question history DB 135 and the feature information related to at least one of the attributes and characteristics of each user included in the user management DB 132.

[0079] FIG. 15 is a diagram showing an example of the feature vector in Modification Example 1. In FIG. 15, in the data row corresponding to the user, in addition to the contents of the question history DB 135, data columns for the school year and the desired school in the user management DB 132 are added. Specifically, data columns corresponding to classifications such as "first-year high school student", "second-year high school student",... for "school year", and classifications such as "University A", "University B",... for "desired school" are added. And for each user, the data column of the corresponding classification is set to "1", and the data column of the non-corresponding classification is set to "null". In this modification example, the feature vector a including the data of the question history DB 135 and the user management DB 132 is used. For elements corresponding to classifications such as "school year" and "desired school", only the classifications where both the target user and the comparison user are "1" are incorporated into the elements of the feature vector, and for classifications where at least one of the target user and the comparison user is "null", they are not incorporated into the elements of the feature vector. According to this method, for a comparison user whose classifications of "school year" and "desired school" match the target user, that is, a comparison user whose attributes and characteristics are common to the target user, the cosine similarity becomes larger and it becomes easier to be specified as an approximate user.

[0080] Note that the feature information of the user can also be used for specifying approximate users without being incorporated into the elements of the feature vector. For example, after calculating the cosine similarity using a feature vector having only the correct / incorrect determination results of the question history DB 135 as elements, a correction may be made to increase the cosine similarity according to the high matching rate of the feature information of the target user and the comparison user. Also, users whose matching rate of feature information is below a predetermined value may be excluded from the extraction target of approximate users.

[0081] <Modification Example 2> Next, Modification Example 2 of the above embodiment will be described. In the above embodiment, a question to be presented is determined from among the words that have not been presented to the target user, according to the predicted correct answer rate. However, even for words that have already been presented, there may be cases where it is meaningful and beneficial to present again the words that the target user answered incorrectly. This is because by presenting again, the learning effect can be confirmed and the learning can be made more firmly established.

[0082] Therefore, in this modification, in addition to the words that have not been presented (unpresented questions), for the words that have been presented to the target user in the past and that the target user answered incorrectly (incorrect answer questions), they are also regarded as "questions to be predicted" and the predicted correct answer rate is calculated. Then, from among the questions to be predicted including the unpresented questions and the incorrect answer questions, according to the predicted correct answer rate, the question to be presented to the target user is determined. Also, in this modification, even words that have been searched for by the target user in the past are subject to being presented. This is because the words of the incorrect answer questions are often searched for for learning after being presented. However, it is not limited to this, and in this modification, words that have been searched for in the past may also be excluded from the questions to be presented. Other points are the same as those in the above embodiment. Below, the differences from the above embodiment will be described. Modification 2 may be combined with Modification 1.

[0083] FIG. 16 is a diagram showing an example of the contents of the question history DB 135 according to Modification 2. Also in this modification, assume that the target user is a user with the user ID "U00000". Also, the approximate users of this target user are, as in the above embodiment, three users with user IDs "U00001" to "U00003" (the cosine similarity thereof is shown in region A). For the words that have not been presented to the target user (in FIG. 16, the words with word IDs "W0002" and "W2995"), the predicted correct answer rate is calculated by the same method as in the above embodiment.

[0084] In this modification example, furthermore, for words that have been presented to the target user in the past and for which the target user has answered incorrectly (that is, words for which the correct / incorrect judgment result is "0"; in FIG. 16, the words with word IDs "W2996" and "W2998"), the predicted correct answer rate is also calculated. For incorrect answer questions as well, based on the data of the correct / incorrect judgment status of approximate users (region B in FIG. 16), the predicted correct answer rate of the target user (region C in FIG. 16) is calculated (derived). That is, the average correct answer rate of approximate users is regarded as the predicted correct answer rate of the target user. Specifically, for the word with word ID "W2996", since 1 out of 3 approximate users answered "1" (correct answer) and 2 answered "0" (incorrect answer), the predicted correct answer rate is calculated as "0.33". Also, for the word with word ID "W2998", since 1 out of 3 approximate users answered "1" (correct answer), 1 answered "0" (incorrect answer), and 1 answered "null" (not presented), the predicted correct answer rate is calculated as "0.50".

[0085] Note that when deriving the predicted correct answer rate for incorrect answer questions, instead of the average correct answer rate of only approximate users, the average correct answer rate of approximate users and the target user may be used as the predicted correct answer rate. In this case, in FIG. 16, for the word with word ID "W2996", out of 4 people including 3 approximate users and the target user, 1 answered "1" (correct answer) and 3 answered "0" (incorrect answer), so the predicted correct answer rate is calculated as "0.25". Also, for the word with word ID "W2998", out of the above 4 people, 1 answered "1" (correct answer), 2 answered "0" (incorrect answer), and 1 answered "null" (not presented), so the predicted correct answer rate is calculated as "0.33".

[0086] Also in this modification example, for all users, the predicted correct answer rate of prediction target questions (not presented questions and incorrect answer questions) is calculated. Further, in this modification example, the predicted correct answer rate calculated for each user is registered in the learning history DB134.

[0087] FIG. 17 is a diagram showing an example of the content of the learning history DB134 according to Modification Example 2. In the learning history DB134 shown in FIG. 17, each user's data block contains data rows (records) corresponding to all words. Also, each data block has a data column (column) for "prediction correct rate", and the calculated prediction correct rate is registered in this data column. By referring to the learning history DB134 in FIG. 17, it is possible to distinguish and obtain the prediction correct rate for unassigned questions and the prediction correct rate for incorrect answers. Specifically, the prediction correct rate in the data row where the data columns of "question date" and "correct / incorrect judgment result" are "null (no data)" can be determined to be the prediction correct rate for unassigned questions. Also, the prediction correct rate in the data row where the "correct / incorrect judgment result" is "0 (incorrect answer)" can be determined to be the prediction correct rate for incorrect answers. Note that for words where the "correct / incorrect judgment result" is "1 (correct answer)", the prediction correct rate is not calculated.

[0088] Note that, similar to the above embodiment, instead of the learning history DB134, the prediction correct rate may be registered in the prediction correct rate DB136. In this case, for example, separately from the first database (the same database as the prediction correct rate DB136 shown in FIG. 12) in which only the prediction correct rate of unassigned questions for each user is registered, a second database in which only the prediction correct rate of incorrect answers for each user is registered may be generated. Thereby, the prediction correct rate for unassigned questions and the prediction correct rate for incorrect answers can be registered in a distinguishable state.

[0089] Next, the prediction correct rate calculation process according to Modification 2 will be described. FIG. 18 is a flowchart showing the control procedure of the prediction correct rate calculation process according to Modification 2. FIG. 18 corresponds to the flowchart of the prediction correct rate calculation process shown in FIG. 13 with step S108 deleted, step S115 added, and step S112 changed to step S112a.

[0090] Steps S101 to S107 in FIG. 18 are the same as steps S101 to S107 in FIG. 13. In step S107, if it is determined that the variable M is less than the total number of words (i.e., "YES" in step S107), the CPU 11 determines whether the word M has been presented to the user (step S109). If it is determined that the word M has been presented to the user (i.e., "YES" in step S109), the CPU 11 refers to the question history DB 135 and determines whether the user N has answered the question of the word M correctly (step S115). If it is determined that the user N has answered the question of the word M correctly (i.e., "YES" in step S115), the CPU 11 determines that the word M is not a prediction target question and sets the prediction correct rate for the word M to "null" (step S111).

[0091] If it is determined that the user N has answered the question of the word M incorrectly (i.e., "NO" in step S115), or if it is determined in step S109 that the word M has not been presented to the user (i.e., "NO" in step S109), the CPU 11 determines that the word M is a prediction target question and calculates the average correct rate of approximate users for the word M as the prediction correct rate for the word M (step S110).

[0092] When step S110 or step S110 ends, the CPU 11 registers the calculation result of the prediction correct rate in the learning history DB 134 (step S112a). As described above, the calculation result of the prediction correct rate may be registered in the prediction correct rate DB 136. The subsequent processing is the same as the flowchart in FIG. 13.

[0093] The flowchart of the test process in this modified example is basically the same as the flowchart of the test process of the above embodiment shown in FIG. 14. However, in the difficulty level specification screen 50 to be displayed in step S201 of FIG. 14, it may be possible to accept the specification of the ratio of presenting unasked questions and wrongly answered questions.

[0094] FIG. 19 is a diagram showing an example of the difficulty level specification screen 50 according to Modified Example 2. The difficulty level specification screen 50 shown in FIG. 19 corresponds to the difficulty level specification screen 50 shown in FIG. 8 with a text box 53 added for specifying the percentage (percent) of the number of questions to be asked for each of the unasked questions and the wrong answer questions. By selecting the test start button 52 with a numerical value from "0" to "100" entered in the text box 53, a plurality of questions are asked so that the number of unasked questions and the number of wrong answer questions to be asked are each at the specified percentage. When a numerical value is entered in one of the text boxes 53 for the unasked questions and the wrong answer questions, a numerical value may be automatically entered in the other text box 53 so that the total of the numerical values in the two text boxes 53 becomes "100".

[0095] Note that on the difficulty level specification screen 50, a difficulty level specification button 51 for unasked questions and a difficulty level specification button 51 for wrong answer questions may be provided separately, and the CPU 11 may be configured to accept the specification of the first difficulty level of unasked questions and the second difficulty level of wrong answer questions, respectively. In this case, the CPU 11 determines, based on the predicted correct answer rates of the unasked questions and the wrong answer questions, the unasked questions corresponding to the specified first difficulty level and the wrong answer questions corresponding to the specified second difficulty level as the questions to be asked to the target user.

[0096] In this modified example, an example in which both unasked questions and wrong answer questions are the subjects of questions has been described, but the present invention is not limited to this, and only wrong answer questions may be the subjects of questions. In this case, only the wrong answer questions may be the target questions for prediction, and only the predicted correct answer rate of the wrong answer questions may be calculated.

[0097] <Effect> As described above, the server 10 as the information processing apparatus according to the present embodiment has a CPU 11 as a processing unit. Based on the question history DB 135 (question history information) including the presence or absence of a question history for a plurality of users regarding a plurality of questions and the correct / incorrect determination results of the questions that have been presented, the CPU 11 identifies, from among the plurality of users, approximate users whose learning tendencies with respect to a certain target user among the plurality of users are similar, and among the plurality of questions, for the target user, a non-questioned question that has not been presented to the target user and a predicted target question that is at least one of the incorrectly answered questions that the target user has answered incorrectly in the past, the predicted correct answer rate of the target user for the predicted target question is derived based on the correct / incorrect determination results of the predicted target questions of the approximate users, and based on the derived predicted correct answer rate, a question to be presented to the target user among the plurality of questions is determined. According to the method of using the correct / incorrect determination results of the approximate users in this way, compared with the prior art in which the predicted correct answer rate is derived from the analysis results such as the degree of understanding of the concepts of the target user himself / herself, the predicted correct answer rate of the target user for the predicted target questions (non-questioned questions and / or incorrectly answered questions) can be derived simply and with high accuracy. Therefore, based on the derived predicted correct answer rate, it is possible to easily present questions of an appropriate difficulty level according to the learning level of the user. For example, in the case of a word test, it becomes possible to predict and present words that the user has not memorized, and efficient improvement of academic ability can be achieved. In addition, by presenting and learning questions with a low predicted correct answer rate, it is possible to practice and acquire the questions that approximate users with a similar learning level have answered incorrectly, so that an advantage can be secured over users who are competitors. In addition, since the approximate users are selected to reflect the academic ability of the target user at that time, for the incorrectly answered questions that the target user has answered incorrectly in the past, it is possible to derive the predicted correct answer rate at that time that reflects the improvement of the academic ability of the target user. Therefore, according to the predicted correct answer rate, the incorrectly answered questions can be presented again at an appropriate timing, and effective learning can be performed for the incorrectly answered questions. In addition, after identifying the approximate users, the predicted correct answer rate of the target user is derived based on the correct / incorrect determination results of the approximate users, so there is no need to analyze the content of the questions for deriving the predicted correct answer rate. Therefore, the above method can be applied to any type of questions.

[0098] Also, the CPU 11 identifies a plurality of approximate users, and sets the average correct answer rate of the prediction target problems of the plurality of approximate users as the prediction correct answer rate. Thereby, the accuracy of the prediction correct answer rate can be improved.

[0099] Also, the CPU 11 identifies approximate users based on the cosine similarity derived from the feature vector including the correct / incorrect determination results of a plurality of problems of the target user and the vector including the correct / incorrect determination results of a plurality of problems of each other user. Thereby, approximate users can be identified by a simple method using the correct / incorrect determination results of a plurality of problems of each user.

[0100] Also, the CPU 11 identifies, as approximate users, users among the plurality of users whose cosine similarity is equal to or greater than a reference value. Thereby, approximate users whose similarity in proficiency tendency with the target user is at a certain level or higher can be identified.

[0101] Also, the CPU 11 identifies, as approximate users, the users of a reference number or reference ratio selected in descending order of cosine similarity among the plurality of users. Thereby, a certain number of approximate users whose similarity in proficiency tendency with the target user is large can be identified.

[0102] Also, the CPU 11 identifies approximate users based on the question history DB 135 and the user management DB 132 (feature information) including at least one of the attributes and characteristics of the plurality of users. Thereby, approximate users can be identified more appropriately, and the accuracy of the prediction correct answer rate of the target user can be further improved.

[0103] Also, the CPU 11 receives a specification of the difficulty level of the questions to be presented to the target user, and determines, based on the prediction correct answer rates of the plurality of prediction target questions, the questions with a prediction correct answer rate corresponding to the specified difficulty level as the questions to be presented to the target user. Thereby, questions with an appropriate difficulty level corresponding to the learning level of the user and the specified difficulty level can be presented.

[0104] Also, in Modification 2, when deriving the predicted correct answer rate for incorrect answer questions, the CPU 11 may use the average correct answer rate of the incorrect answer questions of the approximate user and the target user as the predicted correct answer rate. Incorrect answer questions can be said to be questions that the target user is likely to answer incorrectly. Therefore, by using the average correct answer rate of the approximate user and the target user instead of the average correct answer rate of only the approximate user, the average correct answer rate can be adjusted in a more appropriate direction, that is, in a lower direction. Thereby, the accuracy of the predicted correct answer rate can be further improved.

[0105] Also, in Modification 2, the CPU 11 accepts the specification of the ratio of unasked questions and the ratio of incorrect answer questions in the questions to be presented to the target user, and based on the predicted correct answer rates of a plurality of questions to be predicted, determines the questions to be presented to the target user so that the number of unasked questions and incorrect answer questions to be presented becomes the specified ratio. Thereby, unasked questions and incorrect answer questions can be presented at a ratio desired by the user.

[0106] Also, the information processing method executed by the CPUs 11 and 21 as the computers of the learning support system 1 as the information processing system according to the present embodiment is based on the question history DB 135 (question history information) including the presence or absence of the question history for a plurality of users and the correct / incorrect determination results of the questions that have been asked for a plurality of questions, identifies an approximate user whose learning tendency for a certain target user among the plurality of users is similar to that of a plurality of questions, and for a predicted target question that is at least one of an unasked question that has not been asked to the target user and an incorrect answer question that the target user has answered incorrectly in the past among the plurality of questions, derives the predicted correct answer rate of the target user based on the correct / incorrect determination results of the predicted target questions of the approximate user, and based on the derived predicted correct answer rate, determines the questions to be presented to the target user among the plurality of questions. According to this, compared with the prior art that derives the prediction accuracy rate from the analysis results such as the degree of understanding of the concept of the target user himself / herself, it is possible to simply and highly accurately derive the prediction accuracy rate of the target user for the prediction target problems (unquestioned problems and / or wrongly answered problems). Therefore, based on the derived prediction accuracy rate, it is possible to easily present problems with an appropriate difficulty level according to the learning level of the user. In addition, it is possible to present the wrongly answered problems again at an appropriate timing reflecting the improvement of the academic ability of the target user, and effective learning can be performed for the wrongly answered problems. Further, after identifying approximate users, since the prediction accuracy rate of the target user is derived based on the correct / incorrect determination results of the approximate users, it is not necessary to analyze the content of the problems for deriving the prediction accuracy rate. Therefore, the above method can be applied to any type of problems.

[0107] Also, the server control program 131 as a program according to the present embodiment causes the CPU 11 as a computer provided in the server 10 as an information processing apparatus to perform a process of identifying, from among a plurality of users, approximate users whose learning tendencies for a certain target user among the plurality of users and a plurality of problems are approximate, based on the question history DB 135 (question history information) including the presence or absence of the question history for a plurality of users and the correct / incorrect determination results of the questions that have been presented, a process of deriving the prediction accuracy rate of the target user for prediction target problems, which are at least one of the unquestioned problems that have not been presented to the target user and the wrongly answered problems that the target user has wrongly answered in the past, based on the correct / incorrect determination results of the prediction target problems of the approximate users, and a process of determining, based on the derived prediction accuracy rate, the questions to be presented to the target user among the plurality of problems. According to this, compared with the prior art that derives the prediction accuracy rate from the analysis results such as the degree of understanding of the concept of the target user himself / herself, the prediction accuracy rate of the target user for the prediction target problems (unquestioned problems and / or wrong answer problems) can be derived simply and with high accuracy. Therefore, based on the derived prediction accuracy rate, it is possible to easily present problems with an appropriate difficulty level according to the user's learning level. In addition, wrong answer problems can be presented again at an appropriate timing reflecting the improvement of the target user's academic ability, and effective learning can be performed for the wrong answer problems. Further, after identifying approximate users, the prediction accuracy rate of the target user is derived based on the correct / incorrect judgment results of the approximate users, so there is no need to analyze the content of the problems for deriving the prediction accuracy rate. Therefore, the above method can be applied to any type of problems.

[0108] <Others> Note that the description in the above embodiment is an example of the information processing apparatus, information processing method, and program according to the present invention, and is not limited thereto. For example, part or all of the processing executed by the server 10 in the above embodiment may be executed by the terminal device 20. When the CPU 21 of the terminal device 20 executes the calculation of the prediction accuracy rate and the selection of problems based on the prediction accuracy rate, the terminal device 20 corresponds to the "information processing apparatus".

[0109] Also, although the problems presented to the user are exemplified as problems asking for the translation or spelling of words in a word test, it is not limited thereto, and any problem for which correct / incorrect judgment is possible may be used. Therefore, the problem presentation method of the above embodiment can be applied to any type of problems (for example, memorization problems, written problems, and listening problems using the voice output unit 26 of the terminal device 20, etc.) in any subject (for example, mathematics, Japanese, etc.). Further, it is not limited to the problems in school education, and can also be applied to the problems in qualification tests such as driving schools and the problems presented in quiz games.

[0110] In the above embodiment, an example in which the user specifies the difficulty level of the problem on the difficulty level specification screen 50 has been described. However, the present invention is not limited to this, and the CPU 11 of the server 10 may extract problems according to a predetermined difficulty level setting. For example, words that are highly likely to be unknown to the user (words with a prediction correct answer rate equal to or lower than a predetermined value) may be extracted. Conversely, for the purpose of confirming the user's understanding, words that are highly likely to be known to the user (words with a prediction correct answer rate equal to or higher than a predetermined value) may be extracted. In these cases, the display of the difficulty level specification screen 50 is omitted.

[0111] In the above description, an example in which the HDD and SSD of the storage unit 13 are used as the computer-readable medium of the program according to the present invention has been disclosed. However, the present invention is not limited to this example. As other computer-readable media, information recording media such as flash memory and CD-ROM can be applied. In addition, a carrier wave is also applied to the present invention as a medium for providing the data of the program according to the present invention via a communication line.

[0112] Regarding the detailed configurations and detailed operations of the respective components of the learning support system 1, the server 10, and the terminal device 20 in the above embodiment, it goes without saying that they can be appropriately changed without departing from the spirit of the present invention.

[0113] Although embodiments of the present invention have been described, the scope of the present invention is not limited to the above-described embodiments, and includes the scope of the invention described in the claims and the equivalent scope thereof. The invention described in the claims first attached to the application of this application is appended below. The claim numbers described in the appendix are as in the claims first attached to the application of this application. 〔Appendix〕 <Claim 1> Based on the question history information including the presence or absence of the question history for a plurality of users regarding a plurality of questions and the correct / incorrect determination results of the questions that have been asked, identify an approximate user among the plurality of users whose proficiency tendency with respect to a target user among the plurality of users is approximated to the plurality of questions. Based on the correct / incorrect determination result of the prediction target problem of the approximate user in the question history information, derive the predicted correct answer rate of the target user for the prediction target problem, which is at least one of the unasked questions that have not been asked to the target user and the incorrect answer questions that the target user has answered incorrectly in the past, Comprising a processing unit that determines questions to be presented to the target user among the plurality of questions based on the derived predicted correct answer rate An information processing apparatus characterized by the above. <Claim 2> The processing unit Identifies a plurality of the approximate users, And uses the average correct answer rate of the prediction target problems of the plurality of approximate users as the predicted correct answer rate The information processing apparatus according to claim 1, characterized by the above. <Claim 3> The processing unit identifies the approximate users based on the cosine similarity derived from a vector including the correct / incorrect determination results of the plurality of questions of the target user and a vector including the correct / incorrect determination results of the plurality of questions of each other user. The information processing apparatus according to claim 1 or 2, characterized by the above. <Claim 4> The processing unit identifies, as the approximate users, users among the plurality of users whose cosine similarity is equal to or greater than a reference value. The information processing apparatus according to claim 3, characterized by the above. <Claim 5> The processing unit identifies, as the approximate users, users of a reference number or a reference ratio selected in descending order of the cosine similarity among the plurality of users. The information processing apparatus according to claim 3, characterized by the above. <Claim 6> The processing unit identifies the approximate users based on the question history information and the feature information including at least one of the attributes and characteristics of the plurality of users. The information processing apparatus according to claim 1 or 2, characterized by the above. <Claim 7> The processing unit Receives a specification of the difficulty level of the questions to be presented to the target user, Based on the prediction accuracy rates of the plurality of problems to be predicted, determine, as the problems to be presented to the target user, the problems with prediction accuracy rates corresponding to a specified difficulty level. The information processing apparatus according to claim 1 or 2, characterized in that. <Claim 8> When deriving the prediction accuracy rate for the incorrectly answered problems, the processing unit uses, as the prediction accuracy rate, the average accuracy rate of the incorrectly answered problems of the approximate user and the target user. The information processing apparatus according to claim 1, characterized in that. <Claim 9> The processing unit receives specifications of the ratio of the unasked problems and the ratio of the incorrectly answered problems in the problems to be presented to the target user, Based on the prediction accuracy rates of the plurality of problems to be predicted, determine the problems to be presented to the target user such that the unasked problems and the incorrectly answered problems have the specified ratios. The information processing apparatus according to claim 1 or 2, characterized in that. <Claim 10> An information processing method executed by a computer of an information processing system, the method comprising: Based on the question history information including the presence or absence of the question history for a plurality of users for a plurality of problems and the correct / incorrect determination results of the questions that have been presented, identify, from among the plurality of users, an approximate user whose familiarity tendency with the target user among the plurality of users is similar to that of the plurality of problems, Derive the prediction accuracy rate of the target user for the prediction target problems, which are at least one of the unasked problems that have not been presented to the target user and the incorrectly answered problems that the target user has answered incorrectly in the past, among the plurality of problems, based on the correct / incorrect determination results of the prediction target problems of the approximate user in the question history information, Based on the derived prediction accuracy rate, determine the problems to be presented to the target user among the plurality of problems, Present the determined problems to the target user. An information processing method, characterized in that. <Claim 11> Causing a computer provided in an information processing apparatus to perform a process of specifying, from among the plurality of users, an approximate user whose proficiency tendency with respect to a certain target user among the plurality of users approximates that with respect to the plurality of problems, based on presence or absence of a problem presenting history for the plurality of users with respect to the plurality of problems and problem presenting history information including correct / incorrect determination results of the presented problems; perform a process of deriving a predicted correct answer rate of the target user with respect to a prediction target problem that is at least one of a non-presented problem that has not been presented to the target user and an incorrectly answered problem that the target user has answered incorrectly in the past, based on the correct / incorrect determination result of the prediction target problem of the approximate user in the problem presenting history information; perform a process of determining, based on the derived predicted correct answer rate, a problem to be presented to the target user among the plurality of problems A program characterized by causing the above to be executed.

Explanation of Signs

[0114] 1 Learning support system (information processing system) 10 Server (information processing apparatus) 11 CPU (processing unit) 12 RAM 13 Storage unit 131 Server control program 132 User management DB 133 Dictionary DB 134 Learning history DB 135 Problem presenting history DB 136 Predicted correct answer rate DB 14 Operation unit 15 Display unit 16 Communication unit 17 Bus 20 Terminal device 21 CPU 23 Storage unit 231 Learning app 24 Operation unit 25 Display unit 26 Audio output unit 27 Communication unit 28 Bus 40 Dictionary screen 50 Difficulty Specification Screen 60 Test Screen N Communication Network

Claims

1. Identifying a similar user from among the plurality of users who has a similar tendency to master the plurality of problems to a certain target user based on question history information including whether or not the plurality of problems have been set to the plurality of users and the results of determining whether the questions have been set; Derive a prediction accuracy rate of the target user for a prediction target question that is at least one of an unasked question that has not been posed to the target user and an incorrect question that the target user has answered incorrectly in the past, based on the correct / incorrect judgment result of the prediction target question of the approximate user in the question setting history information; A processing unit is provided for determining a question to be presented to the target user from among the plurality of questions based on the derived prediction accuracy rate.

23. An information processing apparatus comprising:

2. The processing unit includes: Identifying a plurality of said proximate users; The average accuracy rate of the question to be predicted of the plurality of approximate users is defined as the prediction accuracy rate.

2. The information processing apparatus according to claim 1,

3. The information processing device described in claim 1 or 2, characterized in that the processing unit identifies the similar user by a cosine similarity derived based on a vector whose elements include the correct / incorrect judgment results of the multiple questions of the target user and a vector whose elements include the correct / incorrect judgment results of the multiple questions of each other user.

4. The information processing apparatus according to claim 3 , wherein the processing unit specifies, from among the plurality of users, a user whose cosine similarity is equal to or greater than a reference value as the similar user.

5. The information processing apparatus according to claim 3 , wherein the processing unit specifies, as the approximate users, a reference number or a reference ratio of users selected from the plurality of users in descending order of the cosine similarity.

6. The information processing apparatus according to claim 1 , wherein the processing unit identifies the similar users based on the question history information and feature information including at least one of attributes and characteristics of the plurality of users.

7. The processing unit includes: Accepting a specification of the difficulty level of questions to be presented to the target user; Based on the prediction accuracy rates of the plurality of prediction target questions, a question having a prediction accuracy rate corresponding to a specified level of difficulty is determined as a question to be presented to the target user.

3. The information processing apparatus according to claim 1, wherein the information processing apparatus is a computer.

8. The information processing device according to claim 1 , wherein, when deriving the predicted correct answer rate for the incorrectly answered question, the processing unit sets an average correct answer rate for the incorrectly answered question of the approximate user and the target user as the predicted correct answer rate.

9. The processing unit includes: Accepting designation of the proportion of the unassigned questions and the proportion of the incorrectly answered questions to be posed to the target user; Based on the prediction accuracy rates of the plurality of prediction target questions, questions to be presented to the target user are determined so that the number of unasked questions and incorrectly answered questions presented to the target user is set to a specified ratio.

3. The information processing apparatus according to claim 1, wherein the information processing apparatus is a computer.

10. An information processing method executed by a computer of an information processing system, comprising: Identifying a similar user from among the plurality of users who has a similar tendency to master the plurality of problems to a certain target user based on question history information including whether or not the plurality of problems have been set to the plurality of users and the results of determining whether the questions have been set; Derive a prediction accuracy rate of the target user for a prediction target question that is at least one of an unasked question that has not been posed to the target user and an incorrect question that the target user has answered incorrectly in the past, based on the correct / incorrect judgment result of the prediction target question of the approximate user in the question setting history information; determining a question to be posed to the target user from among the plurality of questions based on the derived prediction accuracy rate; Presenting the determined problem to the target user 23. An information processing method comprising:

11. A computer provided in the information processing device A process of identifying a similar user who has a tendency to master the plurality of problems similar to a certain target user among the plurality of users, based on question history information including whether or not the plurality of problems have been set to the plurality of users and a result of determining whether the questions have been set; A process of deriving a prediction accuracy rate of the target user for a prediction target question that is at least one of an unasked question that has not been asked to the target user and an incorrect answer question that the target user has answered incorrectly in the past, based on the correct / incorrect judgment result of the prediction target question of the similar user in the question setting history information; A process of determining a question to be presented to the target user from among the plurality of questions based on the derived prediction accuracy rate. A program characterized by executing the above.

Images