JP5186342B2 - Information processing apparatus, information processing method, and program - Google Patents

Description

  The present invention relates to an information processing technique for learning support, and more particularly to an information processing apparatus, an information processing method, and a program for automatically creating a selection type question in a test question.

  With recent development and popularization of information processing technology, electronic learning (E-learning) using computers and the World Wide Web has become widespread. In addition, the cram school has independently developed a learning support system in order to support learning of knowledge by learners. In such a learning support system in a cram school, a subject person has to create a question for learning knowledge, and the burden on the subject person who is the subject is large. Conventionally, in order to reduce the labor of such a questioner, a question sentence and a question including a correct answer and an incorrect answer are prepared in advance, and a test question is automatically generated from the set of questions. Various learning support systems have been proposed.

  For example, Japanese Patent Application Laid-Open No. 2003-29612 (Patent Document 1) describes a specific symbol in a problem sentence manuscript data file in a computer-based learning system that handles an alternative problem of selecting one correct answer from a plurality of candidates. Extract the enclosed character string, hold the obtained character string array in the memory as a list of answer character strings, leave the character string surrounded by specific symbols in the question text data file as blank, The character string is displayed as a question sentence as it is, and the answer character string candidate list is displayed at the blank position where the learner pays attention. In addition to the correct answer character string, this list includes the answer character string list. A technique is disclosed in which character strings other than correct character strings are randomly extracted and included so that even a beginner using a computer can easily create an alternative question sentence.

  Japanese Patent Laid-Open No. 2004-144903 (Patent Document 2) describes each answer option in a sub-question by a shuffle means based on a question sentence created by a questioner and data obtained by arranging a plurality of answer options. A plurality of types of test questions are created by performing a rearrangement process that changes the arrangement of each sub-question in the arrangement and / or test questions, and these are printed on each question paper by a printing device. A technique is disclosed that is distributed to the respondents and places different types of test questions for each answerer.

In addition, Japanese Patent Laying-Open No. 2005-321768 (Patent Document 3) provides problem information for the purpose of providing a learning system, a learning management server, and a program that can provide learning material information suitable for a user's learning level. A technique for generating an evaluation based on the answer information corresponding to, and introducing one or more e-learning courses from a plurality of e-learning courses based on the evaluation is disclosed.
JP 2003-29612 A JP 2004-144903 A JP 2005-321768 A

  As described above, in order to automatically generate a selection type question, in addition to the question sentence and its correct answer, a plurality of wrong answers must be prepared in advance, which is a burden on the examinee. In the prior art disclosed in Patent Document 1, a character string is extracted from the question sentence manuscript file data and answer candidates including an incorrect answer are generated, so that the burden of creating the answer data of the questioner is somewhat reduced. I can say that.

  However, since the prior art of Patent Document 1 displays answer options that are randomly selected from all the words and phrases specified in the blank, the correct answer of another question is included in the wrong answer of a question. However, this is not sufficient in that the probability of giving hints to learners unintentionally increases. In addition, although it is possible to sort the answer character string list and include the wrong answer with the same correct answer and the first character string in the option, or to include the same number of characters as the correct answer in the option, it is conceptually far This is not sufficient in that it is more likely that an unqualified option will be included in the answer options. Also, the erroneous answers included in the options are limited to the words and phrases specified in the blanks in the question text file data, which is not sufficient from the viewpoint of the variety of options.

  In addition, the prior art disclosed in Patent Document 2 described above includes a plurality of questions in addition to a question sentence and its correct answer in advance in order to shuffle the answer option array and / or the question answer sequence in the question by the shuffle means. It was not possible to reduce the burden on the tester of creating an incorrect answer.

  The present invention has been made in view of the above problems, and in the creation of a test question, the present invention creates choices for answering questions by using the answer results of past test questions by the learner. Provided are an information processing apparatus, an information processing method, and a program for automatically generating a selection-type question that has high diversity and includes an appropriate choice as to a specific question sentence while reducing labor For the purpose.

  In the present invention, in order to solve the above-described problem, the information processing apparatus receives an input of an answer result for the completed test question, and associates the data value of the answer to the question of the completed test question with the question in the database. Store it. In addition, the data value of the answer to the stored question is analyzed, and a statistical value related to the appearance frequency of the data value of each answer in the set of answer data values is calculated. Then, depending on the statistical value, a plurality of options are selected from the set of answer data values for the question, and a selection type question including the plurality of options is created.

  In the above configuration, each option included in the selection type question is selected from data values of erroneous answers that the learner or the like has actually made wrong in the past test questions. For this reason, it is possible to suitably reduce the probability that options that are completely irrelevant to the question to be asked are included. In addition, the above configuration makes it possible to create a selective review question from a descriptive test question conducted on paper or the like.

  In the present invention, the condition for selecting the plurality of options can be changed according to the designated difficulty level. Further, in the present invention, the condition for selecting the plurality of options may include a designation of giving priority to the data value of the answer having the upper or lower frequency corresponding to the difficulty level. In the present invention, the condition for selecting the plurality of options can include specifying the number of options corresponding to the difficulty level.

  In the above configuration, each option is further selected according to the result of statistical analysis processing of the answer results accumulated according to the degree of difficulty. For this reason, it is possible to adjust the difficulty level of questions by preferentially including wrong answers that were easily mistaken for learners in the choices, and preferentially including incorrect answers that were not often mistaken in the choices. Is possible.

  Further, according to the present invention, an index value representing the degree of learning understanding of the answering learner can be associated with the data value of the answer and stored in the database. Then, a statistical value can be calculated for the population for each index value. Further, the statistical value may include a frequency in the population for each of the data values of the answer. According to the above configuration, since the question can be customized according to the learning comprehension degree assigned to the learner, it is possible to create a question that matches the degree of understanding of the learner.

  Furthermore, in the present invention, a learner identification value for identifying a learner who has answered can be stored in the database in association with the data value of the answer. Then, it is possible to select an option by giving priority to the data value of erroneous answers answered in the past by the learner who is the subject of the selection type question. With the above configuration, it is possible to create a customized test question for each learner by utilizing the history information of the learner's test question.

  Moreover, the data value of the said answer can be made into the data value of either the wrong answer by a learner, a correct answer, or these. Further, when giving priority to the data value of the answer having the upper or lower frequency corresponding to the difficulty level, the data value of the wrong answer having the higher frequency and the data value of the correct answer having the lower frequency, or these The priority can be increased by giving priority to either one of the above. Alternatively, the difficulty level can be reduced by giving priority to either the correct answer data value having a higher frequency and the incorrect answer data value having a lower frequency, or one of these.

  Hereinafter, although this invention is demonstrated with embodiment, this invention is not limited to embodiment mentioned later. In the following embodiments of the present invention, a learning support system for supporting learning, including a computer device that creates alternative test questions for measuring a learner's degree of learning understanding, will be described as an example. To do.

  FIG. 1 is a schematic diagram of a learning support system 10 according to the first embodiment of this invention. A learning support system 10 shown in FIG. 1 includes a computer device 12 that creates an alternative test question. The computer device 12 is generally configured as a personal computer or the like.

  The computer device 12 is connected to a database 14 that stores various data for creating alternative test questions. The tester of the alternative test questions inputs the data manually, for example, manually using the computer device 12 while referring to the answer result P from the learner or the like to the already written test questions. The result P is digitized. Data digitized from the answer result P (hereinafter referred to as answer data) includes information that is used as a material for creating a test question. The answer data will be described in detail later. More specifically, the answer data includes a character string or a numerical value of an incorrect answer included in the answer result P (hereinafter referred to as a data value including the character string and the numerical value). The answer data input to the computer device 12 is stored in the database 14 in a predetermined format.

  The computer device 12 shown in FIG. 1 is further connected to the scanner 16 via an interface such as a USB. A test questioner can digitize the answer result P by causing the scanner 16 to sequentially read the answer result P from the learner or the like. The scanner 16 reads the answer result P as an image and transmits the read image data to the computer device 12. The computer device 12 performs image processing on the read image data of the answer result P, and extracts an answer to each question included in the answer result P by, for example, character recognition processing (offline handwritten character recognition processing). At this time, the examiner can appropriately modify the data value obtained by the character recognition.

  Subsequently, the computer device 12 compares the data value of the answer obtained from the answer result P with the data value registered as the correct answer, and automatically scores. Further, the computer device 12 extracts a data value of an incorrect answer used as a material for creating an alternative test question from the data value of the scored answer, and digitizes the answer result P. Similarly, the answer data extracted by the computer device 12 is stored in the database 14 in a predetermined format.

  The computer apparatus 12 performs statistical analysis processing on the answer data stored in the database 14 and creates a new alternative test question according to the obtained statistical analysis result. The computer device 12 shown in FIG. 1 is connected to the printer 18 via an interface such as a USB. The computer device 12 can instruct the printer 18 to output the newly created alternative test question to the printer 18 in response to an instruction from the questioner. The printer 18 prints out the created alternative test question Q in accordance with a command from the computer device 12.

  FIG. 2 shows a functional block 100 of the computer device 12 of this embodiment. Each functional unit (details will be described later) included in the computer device 12 shown in FIG. 2 expands the program on the memory, and controls the operation of each hardware resource by executing the program. Realized.

  The computer device 12 includes a user interface unit including an input device such as a display, a mouse, and a keyboard, and an operation event corresponding to an input to the input device is controlled by each function under the control of an appropriate operating system (OS). The processing result in response to the operation event is output to an output device such as a display.

  The computer device 12 of this embodiment includes an answer input unit 110 to which data digitized from the answer result P is input. FIG. 3A shows an outline of the answer results from the learners and the like for the descriptive test questions that have already been performed. The answer result 200 (P) shown in FIG. 3A includes a class name entry field 202 in which the class name to which the learner belongs is entered, a name entry field 204 in which the name of the learner is entered, and answers to the questions. And each answer entry column 206 in which is entered. For use in creating a test question, for example, a learner's name and an incorrect answer to an incorrect answer question are extracted from the answer result 200. In the learning support system 10, the learner name is used as information for identifying the learner. The information for identifying the learner is not particularly limited, and a student ID number assigned to the learner may be used. Further, the learner of the descriptive test question and the learner who is the subject of the created test question do not need to match.

  Referring to FIG. 2 again, the answer input unit 110 receives the data of the answer result P, and associates the question that was incorrect, the incorrect answer to the question, and the name of the learner who performed the answer. The answer data 140 is stored in the database 14. The answer input unit 110 functions as an input unit in the present embodiment.

  The computer apparatus 12 of the present embodiment further includes an answer analysis unit 120 that analyzes the answer data 140 stored in the database 14. The answer analysis unit 120 functions as an analysis unit in the present embodiment. The answer analysis unit 120 sequentially reads the answer data 140 from the database 14, performs statistical analysis processing, and stores the results of the statistical analysis processing as analysis result data 160 in the database 14. Details of the statistical analysis processing will be described later.

  The answer analysis unit 120 can refer to information about the learner (hereinafter referred to as “learner data”) 150 stored in the database 14 as appropriate. More specifically, the answer analysis unit 120 is an indicator of the degree of learning understanding of the learner (hereinafter referred to as learning comprehension level), such as the name of the class to which the learner belongs, the score of the past exam, and the deviation value of the learner. Can be used for statistical analysis processing.

  FIG. 4A illustrates the data structure of learner data according to the first embodiment of the present invention. The learner data 250 shown in FIG. 4A is a table including a name field 252 in which a name for identifying a learner is input and a class name field 254 in which a class name to which the learner belongs is input. In addition, the learner data 250 may include a field for inputting information on the past test scores, learner deviation values, and the like as appropriate.

  Referring back to FIG. 2, the computer apparatus 12 according to the present embodiment further includes a question creation unit 130. The question creation unit 130 reads the analysis result data 160 and creates a set of questions to be included in the test question according to the test question creation condition R specified by the questioner.

  More specifically, the question creation unit 130, in accordance with the creation condition R, a question sentence stored and registered as the question data 170 in the database 14, and a plurality of questions selected according to the analysis result data for the question sentence. Create a choice question that includes choices. The alternative test question Q including the question created by the question creation unit 130 according to the creation condition R is stored as a file or printed out as described above. The question data 170 holds a question sentence registered in the learning support system 10 together with a question ID for identifying the question.

  FIG. 4B illustrates a data structure of question data according to the first embodiment of the present invention. The question data 260 shown in FIG. 4B includes a question ID field 262 in which a question ID for identifying a question is input, a question sentence field 266 in which the text of the question sentence is input, and a correct answer field in which the correct answer is input. 268. Further, the question data 260 may include a classification field 264 into which a question classification is input. Here, in the classification field 264, a value defined for classifying a question range such as a subject or a unit is input. Although the details will be described later, the classification value is used to select a question in a desired question range from questions registered in the learning support system 10.

  The question format is not particularly limited, and any format such as a question format, a blank supplement format, and a rearrangement format can be adopted. Further, each question may include multimedia data such as an image, sound, or moving image instead of or in addition to the question sentence.

  FIG. 3B shows an outline of the output alternative test questions. An alternative test question 210 shown in FIG. 3B includes a question sentence 212 selected according to the creation condition R, and a plurality of options 214 for the question sentence 212 selected according to the creation condition R and analysis result data. And an answer entry field 216 for entering an answer to each question. The plurality of options include a predetermined number of correct answers and incorrect answers. The number of correct answers and the number of incorrect answers included in the options depend on the creation condition R. In the description of the present embodiment, an alternative set of questions with the number of correct answers of 1 is created. However, the questions can be created with a multiple selection type of questions with a number of correct answers, and is particularly limited. Is not to be done. In addition, you can create a set of questions that ask the number of correct answers.

  Hereinafter, with reference to FIGS. 5 to 7, the details of the method of inputting answer data by the answer input unit 110 will be described. FIG. 5 is a flowchart showing a method for manually inputting answer data according to the first embodiment of the present invention. It should be noted that the flowchart shown in FIG. 5 includes steps performed offline.

  The answer data input method shown in FIG. 5 starts from step S100. In step S101, the examiner first performs a descriptive test on a plurality of learners, and collects the answer result P as shown in FIG. In step S102, the questioner scores the answer result P collected from the learner or the like. In step S103, the questioning person extracts, from each scoring result of the answer result P of the learner or the like, a question that was an incorrect answer and an incorrect answer to the question.

  In step S104, the examiner digitizes the question using the computer device 12 prior to digitizing the answer result P. The computer device 12 receives data input from the questioner and first sets the questions included in the performed descriptive test questions.

  For example, if a question included in the completed descriptive test is not registered in the learning support system 10, the questioner inputs a question sentence and its correct answer, and newly registers the question in the learning support system 10. At this time, a unique question ID for identifying the question is assigned. If the question included in the completed descriptive test is already registered in the learning support system 10, the questioner searches for the question included in the completed descriptive test question from the pre-registered questions. List up and prepare for data input of subsequent answer result P.

  After all the questions are set, for example, a graphical user interface (GUI) for inputting data of the answer result P for each learner is displayed on the display screen of the computer device 12. This GUI includes a list display of questions included in the already described descriptive test questions, a text box for inputting wrong answers to each question by a learner, and the like.

  In step S105, the questioning person digitizes the answer result P using the computer device 12 via the GUI as described above. The computer device 12 receives an incorrect answer question and an input of an incorrect answer for each learner from the questioning person. A data value indicating an invalid answer is also defined in preparation for the case where an invalid answer is included in the answer result P.

  In step S106, the computer device 12 stores the question ID of the question that was incorrect, the wrong answer to the question, and the name of the learner who made the answer as a set in the database 14 as answer data 140. In step S107, the process ends.

  FIG. 6 illustrates a data structure of answer data according to the first embodiment of the present invention. The answer data 220 shown in FIG. 6 includes a name field 222 in which a learner's name is input, and a table in which an incorrect answer to each question is input. More specifically, the table includes a question ID field 226 in which a question ID for identifying each question is input, and an erroneous answer field 228 in which an incorrect answer corresponding to the question ID is input. Note that the answer data 220 may appropriately include a score field 224 into which the score of the learner's test question is input.

  FIG. 7 is a flowchart showing a method for automatically inputting answer data by the scanner according to the first embodiment of the present invention. It should be noted that the flowchart shown in FIG. 7 includes steps performed offline.

  The answer data input method shown in FIG. 7 is started in step S200. In step S <b> 201, first, the examiner uses the computer device 12 to input data of a descriptive test question text. Upon receiving data input from the questioning person, the computer device 12 first sets and registers the question of the descriptive test question. Here, it is assumed that questions and correct answers to be included in the descriptive test questions are registered in the learning support system 10 in advance.

  In step S202, after all questions are set, a descriptive test question is printed out according to a predetermined template. This template defines a position range on the paper where information to be read is written in order to perform subsequent character recognition processing. In accordance with this template, a descriptive test question sheet in which a class name entry field 202, a name entry field 204, and an answer entry field 206 are provided at intended positions is printed.

  In step S203, the questioner performs a descriptive test on a plurality of learners, and collects answer results P as shown in FIG. In step S204, the questioning person causes the scanner 16 to sequentially read a plurality of answer results P such as learners, and performs character recognition processing according to the template. With the character recognition process, for each answer result P, the character strings entered in the class name entry field 202, the name entry field 204, and the answer entry field 206 are extracted, and the data value of the answer to each question is obtained. Note that if the literacy rate of the character recognition process is insufficient, the examiner may appropriately modify the data value obtained by the character recognition.

  In step S205, the data value of the answer to each question obtained and the data value registered as the correct answer are compared and scored, and the question that was incorrect and the wrong answer of the question are extracted. In step S206, the computer apparatus 12 stores the question ID of the question that is an incorrect answer, the wrong answer to the question, and the name of the learner who made the answer as a set in the database 14 as the answer data 140. In step S207, the process ends.

  The details of the statistical analysis processing of the answer data 140 by the answer analysis unit 120 will be described below with reference to FIGS. 8 and 9. FIG. 8 is a flowchart showing statistical analysis processing of answer data according to the first embodiment of the present invention. The statistical analysis processing shown in FIG. 8 is started from step S300, for example, after the answer data 140 is sufficiently accumulated in the database 14 and receives an instruction from the questioning person. In the loop of step S301 and step S306, the processing from step S302 to step S305 is repeated for each question registered in the learning support system 10.

  In step S <b> 302, the computer device 12 reads out the wrong answer by the learner and the learning understanding level of the learner associated with the wrong answer for each question. In step S302, for example, the database 14 is inquired with the question ID to be processed, and a set of erroneous answers having the question ID is acquired. Each element of the set of wrong answers may include a wrong answer data value, a learner name and score associated with the wrong answer data value, a class name associated with the learner name, and the like. The class name or score can be used as a learning comprehension level. In addition, the set of wrong answers to be subjected to statistical processing may be configured to include all the wrong answers obtained so far by the learner or the like, and if sufficient statistical values are obtained, learning is possible. It may be composed of a sample randomly extracted from all wrong answers by the person or the like.

  In step S <b> 303, the computer device 12 performs standardization processing and NG word filtering processing on the data values of the erroneous answers of the elements included in the obtained wrong answer set. In standardization, fluctuations in notation, synonyms, synonyms, etc. are absorbed, and data values are standardized. NG word filtering also excludes ineligible and unpleasant terms from the query set for inclusion in the exam question options.

  In step S304, the computer apparatus 12 counts the elements of the same data value included in the set of wrong answers, and calculates the number of times each wrong answer data value appears, that is, the frequency of each wrong answer data value. At this time, the computer apparatus 12 constructs a population for each learning comprehension level to be associated, and calculates the frequency of the data value of each incorrect answer for each learning comprehension level. The computer device 12 further calculates the frequency of the data value of each wrong answer with the entire learner having no relation to the learning comprehension level as a population. The array of the frequency of the erroneous answer data values is taken as analysis result data 160. This analysis result data 160 is a frequency distribution table summed up according to the whole learner (hereinafter simply referred to as the whole) and the degree of learning comprehension using the data value of each incorrect answer as a nominal scale variable, or the data value of the wrong answer and the learning. It is configured as a contingency table with the degree of understanding as a variable.

  In step S <b> 305, the computer device 12 stores the analysis result data 160 obtained for each question in the database 14. After exiting the loop of step S301 and step S306, the computer apparatus 12 ends the statistical analysis process in step S307.

  FIG. 9 illustrates a data structure of analysis result data according to the first embodiment of the present invention. FIG. 9A shows the data structure of the entire frequency distribution data of a specific question. The overall frequency distribution data 230 shown in FIG. 9A has a question ID field 232 into which a question ID is input, and is created for each question. The overall frequency distribution data 230 further includes a table including an answer field 234 into which an answer is input and a frequency field 236 into which the frequency is input.

  FIG. 9B shows the data structure of frequency distribution data of specific questions and learning comprehension levels. The frequency distribution data 240 of a specific learning comprehension level has a question ID field 242 to which a question ID is input and a learning comprehension level field 244 to which a learning comprehension level is input. Created every time. In this description, class names are adopted as learning comprehension levels. The frequency distribution data 240 of a specific learning comprehension level further includes a table including an answer field 246 into which an answer is input and a frequency field 248 into which the frequency is input.

  The frequency corresponding to each answer input in the frequency field 236 of the overall frequency distribution data 230 shown in FIG. 9A is the frequency over the learning comprehension level for each answer shown in FIG. Corresponds to the total value of. 9A and 9B also show the frequency of correct answer data values in addition to incorrect answers. The frequency of the data value of the correct answer can be calculated, for example, by appropriately calculating the difference between the total number of answers and the incorrect and invalid answers. Further, the frequency can be expressed by an absolute frequency or a relative frequency.

  FIG. 9C is a stacked bar graph illustrating analysis result data aggregated for a specific question. In FIG. 9C, the data value of the mode answer is rearranged in the descending order of frequency. As shown in FIG. 9C, in the present embodiment, the frequency for each incorrect answer data value is obtained for each of the plurality of classes (S, A, B).

  Hereinafter, the question creation processing by the question creation unit 130 will be described in detail with reference to FIGS. FIG. 10 is a flowchart showing a question creation process according to the first embodiment of the present invention. The question creation process shown in FIG. 10 is started from step S400 in response to an instruction from the questioner after the statistical analysis process for the answer data 140 is completed, for example.

  In step S <b> 401, first, the questioning person inputs data for the test question creation condition R using the computer device 12. The computer device 12 receives the data input from the questioner and determines the creation condition R.

  The creation condition R inputted by the data from the questioner includes, for example, designation of difficulty and condition designation for selecting a question to be asked. The questioner can specify that questions are automatically selected from a specific range of questions in the registered questions. Alternatively, the examiner can specify that a specific question is directly selected. When automatic selection is designated, the creation condition R can include designation of a category indicating the question range and designation of the number of questions in each question range. In the case where manual selection of questions is designated, the creation condition R can include designation of specific questions.

  The difficulty level affects the tendency to select options for questions. In the learning support system 10, for each difficulty level that can be specified, the number of options prepared for the question, the population based on the frequency used when selecting the option, and the tendency to select the option are specified. Etc. are defined. The difficulty level that can be specified is registered in advance in the learning support system 10 or is user-defined by the examiner. The difficulty level can be specified for the entire test question to be created, and can be specified for each question of the test question, and is not particularly limited.

  The designation of the population designates whether to use a statistical analysis result of the entire learner or a statistical analysis result based on a population having a specific learning understanding level. Specifying the tendency to select the above choices specifies whether to give priority to erroneous answers that are frequently mistaken and to be included in choices, or to give priority to incorrect answers that are less likely to be mistaken. To do.

  In step S <b> 402, the computer device 12 reads out data related to the question selected as a question in accordance with the creation condition R. The data related to the question includes analysis result data 160 and a question sentence of question data 170. When automatic selection is designated, questions to be asked are selected from questions in the designated question range at random or according to a predetermined rule.

  In the loop of step S403 and step S408, the processing from step S404 to step S407 is repeated for each question selected as a question.

  In step S404, the computer apparatus 12 determines the number of options according to the creation condition R. The number of question options is defined in association with the difficulty level. The difficulty level is defined such that, for example, the number of options increases as the difficulty level becomes higher than the standard, and the number of options decreases as the difficulty level becomes lower than the standard.

  In step S <b> 405, the computer apparatus 12 determines a population based on the frequency for selecting an option according to the creation condition R. The designation of the population as the basis of the frequency used when selecting an option is also defined in association with the difficulty level. In this determination, it is determined whether to use the frequency distribution data of the entire learner regardless of the learning comprehension level or to use the frequency distribution data of a specific learning comprehension level. When using the frequency distribution data of a specific learning comprehension level, it is further defined which of the learning comprehension level frequency distribution data is used. The difficulty is, for example, a degree based on a population with higher learning comprehension as the difficulty becomes higher than the standard, and the comprehension becomes lower as the specified difficulty becomes lower than the standard. It is defined to use the frequency based on the population of learning comprehension.

  In step S406, the computer device 12 uses the frequency distribution data based on the determined population to calculate the number of erroneous answers necessary for the options (the number of correct answers from the number of options determined in step S404 (that is, the questions of alternatives). Then, the number obtained by subtracting 1): hereinafter, referred to as the required number of incorrect answers) is selected as an option. The details of the method for selecting options will be described later.

  In step S <b> 407, the computer device 12 determines the selected wrong answer and the number of correct answer options and the order of description. For example, the numbers and description order are shuffled at random in order to eliminate the bias of the questions.

  After exiting the loop of step S403 and step S408, in step S409, the computer apparatus 12 creates a test question document file including a question sentence and a question to be asked consisting of a plurality of options using a template. Then, the computer device 12 issues a print output command for the designated number of copies of the created test question to the printer 18 and terminates the question creation processing in step S410. In addition to the test questions, a document file of the model questions and answers is also created.

  Hereinafter, the details of the method of selecting options for a question will be described. FIG. 11 is a diagram illustrating an option selection method for a question, together with analysis result data to be used and a test question to be output. FIG. 11A is a diagram illustrating an example of an option selection method using the entire frequency distribution data. In the example shown in FIG. 11A, the analysis result data as a basis for selecting an option is the entire frequency distribution data as shown in FIG. 9A.

  In the option selection method shown in FIG. 11 (A), among the erroneous answer data values, those having higher frequencies in the overall frequency distribution are selected with priority. In other words, it can be said that the difficulty level is higher because more learners preferentially select the wrong answer. Here, selecting with priority means selecting an erroneous answer data value with a probability corresponding to the frequency or selecting a data value from a limited range. For example, a value obtained by weighting the relative frequency (0 to 1 or 0% to 100%) in the entire incorrect answer excluding the correct answer according to the difficulty level can be used as the probability corresponding to the frequency.

  In addition, the range of the cumulative relative frequency accumulated from the altitude number side except for the correct answer can be defined in association with the difficulty level. As a range designation method, for example, the cumulative relative frequency is designated in the form of “0% to 70%” from the altitude number side. In the example shown in FIG. 11 (A), the cumulative relative frequency of data values up to “Cash” is approximately 67% (excluding correct answers), and “Ao” and “Cashoku” are in the range 292 for selecting an incorrect answer. It becomes. In FIG. 11A, three of “Ao” and “Cashoku” plus “Ao Murasaki” as the correct answer are selected as options 274 of “Question 1” 272.

  If the number of elements included in the obtained range is less than the required number of incorrect answers (“2” in the example of FIG. 11), the range should be expanded from the high frequency side until the required number is reached. Good. In addition, when elements exceeding the required number of erroneous answers are included, the elements can be selected randomly or according to a predetermined probability. In addition, the range may be specified in the form of a range of necessary erroneous answers from the altitude side.

  FIG. 11B is a diagram illustrating another example of an option selection method using the entire frequency distribution data. In the example shown in FIG. 11B, as in FIG. 11A, the wrong answer options are selected based on the entire frequency distribution data shown in FIG. 9A. However, in the option selection method shown in FIG. 11B, among the erroneous answer data values, those having lower frequencies in the overall frequency distribution are selected preferentially. In other words, it can be said that the degree of difficulty is relatively low because the wrong answer of the minority group is selected with priority. In this selection method, the relative frequency (0 to 1 or 0% to 100%) of all the incorrect answers excluding the correct answer is subtracted from the total (1 or 100%) (that is, 1 to 0 or 100% to 0%). The weighted value according to the difficulty level can be set as the probability according to the frequency.

  In addition, the range of the cumulative relative frequency accumulated from the low frequency side excluding the correct answer can be defined in association with the difficulty level. For example, the cumulative relative frequency from the low frequency side is specified in a format such as “1% to 30%”. In the example shown in FIG. 11B, the cumulative relative frequency of data values up to “Aka” is about 33% (excluding correct answers), and up to “green” is 12% (excluding correct answers). “Kiro” is the range 296 for selecting the wrong answer. If the number of elements included in the obtained range is less than the required number of incorrect answers, similarly, the range may be corrected from the low frequency side until the required number is reached. In addition, when elements exceeding the necessary number of wrong answers are included, they can be selected at random.

  In the example shown in FIG. 11B, “green” and “black” which are elements within the range are selected as they are, and the correct answer “Aomura Saki” is added to “question 1” 282. Options 284 are obtained. In addition, as a range designation method, there is a case where designation is made in the form of a range of necessary erroneous answers from the low frequency side.

  FIG. 12 is a diagram showing another method for selecting options for a question, together with analysis result data to be used and a test question to be output. FIG. 12A is a diagram illustrating an example of an option selection method using frequency distribution data of a specific learning comprehension level. In the example shown in FIG. 12A, the analysis result data as a basis for selecting an option is frequency distribution data of a specific learning comprehension degree as shown in FIG. 9B.

  In the option selection method shown in FIG. 12 (A), among the data values of erroneous answers, those having higher frequencies in the frequency distribution obtained for a specific learning comprehension level are selected preferentially. That is, there are many learners with a specific learning comprehension level, and the wrong answer is selected with priority. For this reason, it can be said that it is a question including choices suitable for the degree of understanding of the learner, such as making a mistake that a learner with a specific learning comprehension level often makes.

  In FIG. 12A, the frequency distribution data obtained for the learning comprehension level of the “S” class is used, and the number of options is increased compared to that shown in FIG. Here, it is assumed that the “S” class is allocated to learners with a higher level of learning understanding. That is, the question shown in FIG. 12 (A) is more difficult because there are many choices, and since the wrong answer that is likely to be selected by a learner with higher understanding is preferentially selected. It can be said.

  FIG. 12B is a diagram illustrating another example of an option selection method using frequency distribution data of a specific learning comprehension level. In the example shown in FIG. 12B, the wrong answer options are selected based on the frequency distribution data of the specific learning comprehension degree shown in FIG. 9B, as in FIG. 12A. However, in the option selection method shown in FIG. 12B, the frequency distribution data obtained for the learning comprehension degree of the “B” class assigned to the learner with a higher degree of learning comprehension is used, and the data value of the wrong answer Among them, the one having the lower frequency in the frequency distribution obtained for the learning comprehension level “B” is preferentially selected. It can be said that the difficulty level is low because the minority's wrong answers are selected with priority. Therefore, it is possible to create a question that increases the motivation for learning of a learner with a low level of understanding of learning.

  In the learning support system 10 according to the first embodiment described above, it is possible to automatically create a new selective question by using an incorrect answer included in the answer result of the descriptive test question that has been performed. Each option is selected from the wrong answers that the learner may have actually made in the past. For this reason, it is possible to suitably reduce the probability that options that are completely unrelated to the question are included.

  Further, each option is selected according to the result of statistical analysis processing of the answer result. For this reason, it becomes possible to adjust the difficulty level of a question by including an incorrect answer that is easily mistaken in an option, or including an incorrect answer that is not often mistaken in an option. Furthermore, since the questions can be customized according to the learning comprehension degree assigned to the learner, it is possible to create a question suitable for the degree of understanding of the learner.

  In the above-described embodiment, the frequency distribution data by class is used as the frequency distribution data by learning comprehension level. However, the index value indicating the degree of learning understanding to be used is not particularly limited. In other embodiments, it is related to the score of the exam that made the wrong answer, associated with the data of the wrong answer, and further related to the score of the past examination question associated with the learner or the deviation value of the learner. The information to be used may be the learning comprehension level, and the frequency distribution data may be created for each range of the score and the deviation value.

  Further, in the above-described embodiment, it has been described that erroneous answer data is created from the answer result P to the descriptive test question. However, in order to increase the number of samples from which the frequency is calculated, for example, the answer data 140 can be generated from the answer results for the generated selective test questions and fed back to each frequency.

  Furthermore, in the above-described embodiment, it has been described that an incorrect answer included in the answer result is extracted, subjected to statistical analysis processing, and an option for an incorrect answer is selected from past incorrect answers. However, for example, in the case of questions that may have multiple answers, a statistical analysis process is performed to count the correct answers answered by the learner, and the correct answer options are selected from the past correct answers according to the frequency. It can also be configured. In this case, it is possible to increase the difficulty level by including a correct answer with a small number of answers in the option, or to decrease the difficulty level by including the correct answer with a large number of answers in the option.

  The learning support system according to the second embodiment will be described below. Note that the learning support system of the second embodiment shares the main parts with the learning support system of the first embodiment, and therefore, the differences will be mainly described. FIG. 13 is a schematic diagram of a learning support system 30 according to the second embodiment of this invention. A learning support system 30 shown in FIG. 13 is configured as a system that supports the learning of a learner online via a network 32 such as the Internet.

  The learning support system 30 includes a server 34 that creates alternative test questions. The server 34 distributes the created alternative test questions to a client computer (hereinafter referred to as a “learner terminal”) 42 used by a learner or the like via the network 32. The server 34 includes the functional blocks shown in FIG. 2 and corresponds to the computer device 12 in the first embodiment.

  The server 34 is connected to a database 36 that stores answer data 140, analysis result data 160, question data 170, and learner data 150 for creating alternative test questions. A tester of an alternative test question uses a client computer (hereinafter referred to as a tester terminal) used by the tester while referring to an answer result P from a learner or the like with respect to an already-explained test question. For example, data is input manually and the answer result P is digitized. Alternatively, the questioner terminal 38 may digitize the answer result P by causing the scanner 16 to sequentially read the answer result P from the learner or the like using the scanner 40 connected via an interface such as USB. it can.

  As described in the first embodiment, the answer data 140 digitized from the answer result P includes an erroneous answer data value included in the answer result P. The answer data 140 extracted by the questioner terminal 38 is transmitted to the server 34. The server 34 receives the transmitted answer data 140 and stores it in the database 36 in a predetermined format. Further, the server 34 performs statistical analysis processing on the answer data 140 accumulated in the database 36 and stores the obtained statistical analysis results in the database 36.

  On the other hand, the learner uses the learner terminal 42 to access the server 34 via the network 32, logs in with the learner's name or the like, and requests distribution of the test questions. The server 34 receives a request for distribution of the test questions and creates a new alternative test question according to the analysis result data 160 stored in the database 36. Then, the server 34 transmits the created alternative test question data to the learner terminal 42 as a response to the test question delivery request.

  The learner uses the learner terminal 42 to transmit to the server 34 the answer results for the delivered alternative test questions. Further, the server 34 can store the answer results for the received alternative test questions in the database 36 as the answer data 140 for use in the statistical analysis process. Note that the functional configuration of the server 34 is the same as that of the computer device 12, and thus detailed description thereof is omitted.

  In the learning support system 30 of the second embodiment, the learner can determine the creation condition R described in the first embodiment. For this reason, the learner can challenge a test question of a desired difficulty level.

  Furthermore, in the learning support system 30 of the second embodiment, the answer result to the alternative test question transmitted from the learner can be reflected in the statistical analysis process. Furthermore, in the learning support system 30 of the second embodiment, when creating an alternative test question, it is possible to create a test question individually for each learner. The answer data is associated with the name of the learner who identifies the learner from the login information, for example, to increase the probability that the learner may have made a mistake in the past when selecting an option. Thus, a learner who requests delivery of a test question can preferentially include a data value of an erroneous answer that has been wrong in the past. With this configuration, it is possible to create a test question customized for each learner by utilizing the history information of the learner's test question.

  Hereinafter, the hardware configuration of the computer apparatus 12 will be described. Since the computer device 12 and the server 34 can adopt the same hardware configuration, only the computer device 12 will be described below.

  FIG. 14 shows a schematic hardware configuration of the computer apparatus 12. The computer device 12 shown in FIG. 14 is capable of processing by a central processing unit (CPU) 52, a cache memory 54 having levels such as L1 and L2 that enable high-speed access of data used by the CPU 52, and the CPU 52. And a system memory 56 formed from a solid-state memory device such as a RAM or a DRAM.

  The CPU 52, cache memory 54, and system memory 56 are connected to other devices or drivers, such as a graphics driver 60 and a network interface card (NIC) 62, via a system bus 58. Yes. The graphics driver 60 is connected to an output device such as an external display 64 via a bus, and can display a processing result by the CPU 52 on a display screen. The NIC 62 connects the computer apparatus 12 to a network using an appropriate communication protocol such as TCP / IP at the physical layer level and the data link layer level.

  An I / O bus bridge 66 is further connected to the system bus 58. A hard disk 70 is connected to the downstream side of the I / O bus bridge 66 by IDE, ATA, ATAPI, serial ATA, SCSI, USB, or the like via an I / O bus 68 such as PCI. An input device 72 such as a keyboard and a pointing device such as a mouse is connected to the I / O bus 68 via a bus such as a USB, and a user interface is provided by the input device 72.

  As the CPU 52 of the computer apparatus 12, any single core processor or multi-core processor can be used, and more specifically, for example, Xeon (registered trademark), Itanium (registered trademark), POWER PC (registered trademark), or the like. A CISC or RISC chip can be mentioned. The computer device 12 is controlled by an OS such as WINDOWS (registered trademark) 200X, UNIX (registered trademark), or LINUX (registered trademark), and develops the program on the system memory 56 under the management of the OS. By executing the operation control of each hardware resource, the configuration and processing of the functional unit described above can be realized on the computer.

  As described above, according to the present embodiment, in the creation of the test questions, by using the answer result of the past test questions by the learner, it is possible to reduce the effort of creating the answer options of the questioner, It is possible to provide an information processing apparatus, an information processing method, and a program that automatically generate a selection-type question that has high diversity and includes appropriate choices for a specific question sentence.

  The information processing apparatus can be implemented with various computer system configurations including a multiprocessor system in addition to the computer apparatus 12 and the server 34 described above.

  The above-described functions of the present invention are described in object-oriented programming languages such as C ++, Java (registered trademark), Java (registered trademark) Beans, Java (registered trademark) Applet, Java (registered trademark) Script, Perl, and Ruby. The program can be realized by a device-executable program, and can be stored in a device-readable recording medium and distributed or transmitted and distributed.

  Although the present invention has been described with specific embodiments, the present invention is not limited to the embodiments, and other embodiments, additions, changes, deletions, and the like can be conceived by those skilled in the art. It can be changed within the range, and any embodiment is included in the scope of the present invention as long as the effects and effects of the present invention are exhibited.

1 is a schematic diagram of a learning support system 10 according to a first embodiment of this invention. The functional block diagram of the computer apparatus of this embodiment. (A) The figure which shows the outline of the answer result from the learner etc. with respect to the description type | formula test question already implemented, and the (B) output alternative type test question. The data structure of the learner data and question data by 1st Embodiment of this invention is illustrated. The flowchart which shows the manual input method of the answer data in 1st Embodiment of this invention. The figure which illustrates the data structure of the answer data by a 1st embodiment of the present invention. The flowchart which shows the automatic input method of the answer data by the scanner in 1st Embodiment of this invention. The flowchart which shows the statistical analysis process of the answer data by 1st Embodiment of this invention. The figure which illustrates the data structure of the analysis result data by a 1st embodiment of the present invention. The flowchart which shows the question preparation process by 1st Embodiment of this invention. The figure which shows the choice selection method with respect to a question with the analysis result data to be used and the test question output. The figure which shows the other selection method of the choice with respect to a question with the analysis result data to be used and the test question output. The figure which shows the schematic of the learning assistance system of the 2nd Embodiment of this invention. The figure which shows the schematic hardware constitutions of a computer apparatus.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 ... Learning support system, 12 ... Computer apparatus, 14 ... Database, 16 ... Scanner, 18 ... Printer, 30 ... Learning support system, 32 ... Network, 34 ... Server, 36 ... Database, 38 ... Questioner terminal, 40 ... Scanner 42 ... Learner terminal, 52 ... CPU, 54 ... Cache memory, 56 ... System memory, 58 ... System bus, 60 ... Graphics driver, 62 ... NIC, 64 ... Display, 66 ... I / O bus Bridge 68 68 I / O bus 70 Hard disk 72 Input device 100 Functional block 110 Answer input section 120 Answer analysis section 130 Question creation section 140 Answer data 150 Learning Data, 160 ... analysis result data, 170 ... question data, 200 ... answer result, 202 ... class Entry column, 204 ... Name entry column, 206 ... Answer entry column, 210 ... Choice test question, 212 ... Question sentence, 214 ... Multiple choices, 216 ... Answer entry column, 220 ... Answer data, 222 ... Learner name field 224 ... Point field, 226 ... Question ID, 228 ... Error answer field, 230 ... Whole frequency distribution data, 232 ... Question ID field, 234 ... Answer field, 236 ... Frequency field, 240 ... Frequency distribution data of learning comprehension 242 ... Question ID field, 244 ... Learning comprehension field, 246 ... Answer field, 248 ... Frequency field, 250 ... Learner data, 252 ... Name field, 254 ... Class name field, 260 ... Question data, 262 ... Question ID Field, 264 ... Classification field, 266 ... Question sentence field 268 ... correct answer field, 272, 282 ... questions 1, 274, 284 ... choices, 292 ... range for selecting wrong answers, 296 ... range for selecting wrong answers, P ... answer results, Q ... choice test questions, R ... Creation conditions

Claims (14)

An information processing apparatus for creating a selective question, wherein the information processing apparatus includes:
An input unit that receives an answer result for the completed test question and stores a data value of the answer to the question of the completed test question in a database in association with the question;
Analyzing the data value of the answer to the stored question and calculating a statistical value related to the frequency of appearance of the data value of each answer in the set of data values of the answer;
An information processing apparatus comprising: a creation unit that selects a plurality of choices from a set of data values of the answer depending on the statistical value and creates a selection-type question including the plurality of choices for the question .   The information processing apparatus according to claim 1, wherein the creation unit changes a condition for selecting the plurality of options according to a designated difficulty level.   The input unit further associates an index value representing the degree of learning understanding of the answered learner with the data value of the answer and stores the index value in the database, and the analysis unit is configured for the index-specific population. The information processing apparatus according to claim 2, wherein the statistical value is calculated.   The information processing apparatus according to claim 3, wherein the statistical value includes a frequency in the population with respect to each data value of the answer.   The information processing apparatus according to claim 4, wherein the condition for selecting the plurality of options includes designation to prioritize a data value of an answer having a higher or lower frequency corresponding to the difficulty level.   The information processing apparatus according to claim 5, wherein the condition for selecting the plurality of options further includes designation of the number of options corresponding to the difficulty level.   The input unit further associates a learner identification value for identifying a learner who has answered with the data value of the answer and stores it in the database, and the creation unit learns to be a subject of the selection type question. The information processing apparatus according to claim 6, wherein a data value of erroneous answers answered by a person in the past is preferentially selected. A method for creating a selective question, the information processing apparatus comprising:
Receiving an answer result for the completed exam question;
Storing data values of answers to the questions of the completed test questions in a database in association with the questions;
Analyzing the data value of the answer to the stored question and calculating a statistical value related to the frequency of appearance of the data value of each answer in the set of data values of the answer;
Selecting a plurality of options from the set of data values of the answer depending on the statistical value, and creating a selection-type question including the plurality of options for the question. .   The information processing method according to claim 8, wherein the creating step includes a sub-step of selecting the plurality of options according to a designated difficulty level.   The storing step includes a sub-step of storing in the database the index value representing the degree of learning understanding of the answering learner in association with the data value of the answer, and the calculating step includes the index The information processing method according to claim 9, further comprising a sub-step of calculating a statistical value for a population by value.   The storing step further includes a sub-step of associating a data value of the answer with a learner identification value for identifying the answering learner, and storing it in the database. The information processing method according to claim 10, further comprising a sub-step of preferentially selecting data values of erroneous answers answered in the past by a learner to be asked questions. A computer-executable program for creating a selective question, comprising:
An input unit that receives an answer result for the completed test question and stores a data value of the answer to the question of the completed test question in a database in association with the question;
Analyzing the data value of the answer to the stored question, and calculating a statistical value related to the frequency of appearance of the data value of each answer in the set of data values of the answer; and On the other hand, a creation unit that selects a plurality of choices from a set of data values of the answer depending on the statistical value, and creates a selection type question including the plurality of choices
A computer-executable program that functions as a computer. An information processing apparatus for creating a selective question, wherein the information processing apparatus includes:
In response to the input of the answer result for the completed test question, the data value of the answer to the question of the completed test question, the learner identification value for identifying the question and the answerer, and the learning of the answerer An input unit that is stored in the database in association with an index value representing the degree of understanding;
Analyzing the data value of the answer to the stored question, and for a population by the index value, related to the set of data values of the answer, and within the population for each of the data values of the answer An analysis unit for calculating a statistical value including the frequency,
Prior to the question, the data value of the answer having the upper or lower frequency corresponding to the designated difficulty is prioritized from the set of data values of the answer, and is subject to the selective question. An information processing apparatus including: a creation unit that prioritizes data values of erroneous answers answered in the past by a learner, selects a number of options corresponding to the difficulty level, and creates a selection-type question including the plurality of options . An information processing apparatus for creating a selective question, wherein the information processing apparatus includes:
An input unit that receives an answer result for the completed test question and stores a data value of the answer to the question of the completed test question in a database in association with the question;
Analyzing the data value of the answer to the stored question and calculating a statistical value related to the frequency of appearance of the data value of each answer in the set of data values of the answer;
A selection unit that selects a plurality of choices from a set of data values of the answer depending on the statistical value, and creates a selection type question including the plurality of choices.
The creation unit changes a condition for selecting the plurality of options according to a designated difficulty level,
The input unit further associates an index value representing the degree of learning understanding of the answered learner with the data value of the answer and stores the index value in the database, and the analysis unit is configured for the index-specific population. Calculate statistics,
The statistical value includes the frequency within the population for each of the data values of the answer,
The condition for selecting the plurality of options includes a designation to prioritize a data value of an answer having a higher or lower frequency corresponding to the difficulty level,
The condition for selecting the plurality of options further includes designation of the number of options corresponding to the difficulty level,
The input unit further associates a learner identification value for identifying a learner who has answered with the data value of the answer and stores it in the database, and the creation unit learns to be a subject of the selection type question. An information processing apparatus that preferentially selects data values of wrong answers answered by a person in the past.

Images