Personalized test question generation method and system

Abstract

The invention relates to the technical field of test question generation, and particularly discloses a personalized test question generation method and system, and the method comprises the steps: obtaining a capability information file of a user, carrying out the recognition of the capability information file, and generating a static capability keyword; acquiring an experience information file of a user, identifying the experience information file, and generating a dynamic capability keyword; selecting test questions from a preset test question bank according to the static capability keywords and the dynamic capability keywords, and synchronously determining the push probability of the test questions; and pushing the test questions based on the pushing probability, obtaining reply information of the user, generating an evaluation value according to the reply information, and synchronously adjusting the pushing process. The keyword is determined according to the information file of the user, the test questions are matched in the question bank according to the keyword, the reply information of the user is obtained, the test question matching process is updated in real time according to the reply information, and the integrating degree of the test questions and the user is greatly improved.

Description

Technical Field

[0001] The present invention relates to the technical field of test question generation, and in particular to a method and system for generating personalized test questions. Background Art

[0002] When it is necessary to test a user, testing with test questions is a relatively common and simple method. The existing testing methods mostly pre-build a question bank. When a test request is received, some test questions are randomly selected and combined to test the user. This method has good randomness, but the fit with the user is not high, which makes the testing process likely to be more accidental and it may be difficult to obtain the user's true situation. Therefore, how to improve the fit between the test questions and the user during the test question selection process is the technical problem that the technical solution of the present invention wants to solve. Summary of the Invention

[0003] The purpose of the present invention is to provide a personalized test question generation method and system to solve the problems raised in the above background technology.

[0004] To achieve the above object, the present invention provides the following technical solutions: A method for generating personalized test questions, the method comprising: Obtaining a user's capability information file, identifying the capability information file, and generating static capability keywords; Obtaining a user's experience information file, identifying the experience information file, and generating dynamic capability keywords; Select questions from the preset question bank based on static and dynamic ability keywords, and simultaneously determine the probability of pushing the questions; Push questions based on the push probability, obtain the user's answer information, generate an evaluation value based on the answer information, and synchronously adjust the push process; Among them, the dynamic ability keywords all contain time tags. In the process of selecting test questions, the application process of the dynamic ability keywords introduces weights based on time tags.

[0005] As a further solution of the present invention, the steps of obtaining the user's capability information file, identifying the capability information file, and generating static capability keywords include: Obtain the educational experience information sent by the user, match similar users based on the educational experience information, and query the educational experience information of similar users within a preset time range as reference educational information; Calculate the similarity between the current user's education experience information and the reference education information; Obtaining the average of the calculated similarities, and determining that the recognition is successful when the average reaches a preset average threshold; The educational experience information is word-located according to the preset keyword table to obtain static ability keywords.

[0006] As a further solution of the present invention, the steps of obtaining the user's experience information file, identifying the experience information file, and generating dynamic capability keywords include: Receive the experience information file input by the user, identify the experience information file, and divide it into project files with time tags; Obtain a description file of any project file, identify the description file, determine the project difficulty, and determine the project radius based on the project difficulty; Acquiring capability requirements of a project file, determining display parameters according to the capability requirements, and constructing experience units of the project file according to the project radius and the display parameters; Count the experience units of the same project file to obtain the experience characteristics; Count all experience features to obtain an experience unit set, select feature units from the experience unit set, obtain the project file corresponding to the feature unit, identify the project file, and obtain dynamic capability keywords; The process of selecting the characteristic unit is as follows: Calculate the difference in display parameters of each pixel point in any two experience units; The parameter differences are cumulatively displayed as unit distances; Calculate the mean unit distance between each experience unit and other experience units, arrange the experience units in increasing order of the mean unit distance, and select a preset number of experience units in sequence as feature units.

[0007] As a further solution of the present invention: the step of selecting a question from a preset question bank based on the static ability keywords and the dynamic ability keywords and simultaneously determining the push probability of the question includes: Set the modifiers for static ability keywords to default values; Obtaining a time tag of a dynamic capability keyword, and determining a correction coefficient according to a radius, display parameters, and time tag of a feature unit of a project file corresponding to the dynamic capability keyword; Calculate the matching degree between each question in the preset question bank and the static ability keywords and dynamic ability keywords; Adjust the matching degree according to the correction coefficient to determine the comprehensive matching degree of the test questions; Determining the probability of pushing the test question based on the comprehensive matching degree; The process of determining the comprehensive matching degree is as follows: Where, For comprehensive matching, is the default value of the correction factor. is the total number of static capability keywords, For the The matching degree of static ability keywords; is the total number of dynamic capability keywords, is the adjustment constant of the dynamic capability keyword, For the The matching degree of static ability keywords, For the Display parameters of static ability keywords, For standard display parameters, For the The time difference between the time tag of a static ability keyword and the current moment, For the As a further solution of the present invention: the steps of pushing questions based on push probability, obtaining user response information, generating evaluation values based on the response information, and synchronously adjusting the push process include: Randomly push test questions based on the push probability until the test questions meet the preset conditions; the condition is that the test questions of different difficulty levels reach the preset number; Obtain the user's response information for each test question, verify the response information based on the reference answer, and generate an evaluation value; Adjust the number of test questions of different difficulty levels according to the evaluation value.

[0008] As a further solution of the present invention: the content of the synchronous adjustment push process also includes: Read the evaluation value of each test question; When the evaluation value is less than the preset evaluation threshold, query the dynamic ability keyword corresponding to the test question; Query the project files and experience units corresponding to dynamic capability keywords; The experience unit is updated according to the evaluation value, and the generation process of the feature unit is updated synchronously.

[0009] The technical solution of the present invention also provides a personalized test question generation system, the system comprising: A static keyword generation module is used to obtain a user's capability information file, identify the capability information file, and generate static capability keywords; A dynamic word generation module is used to obtain the user's experience information file, identify the experience information file, and generate dynamic ability keywords; A push probability determination module is used to select test questions from a preset test question bank based on static ability keywords and dynamic ability keywords, and simultaneously determine the push probability of the test questions; A negative feedback adjustment module is used to push test questions based on the push probability, obtain the user's response information, generate an evaluation value based on the response information, and synchronously adjust the push process; Among them, the dynamic ability keywords all contain time tags. In the process of selecting test questions, the application process of the dynamic ability keywords introduces weights based on time tags.

[0010] As a further solution of the present invention: the static word generation module includes: A similar user matching unit is used to obtain the educational experience information sent by the user, match similar users based on the educational experience information, and query the educational experience information of similar users within a preset time range as reference educational information; A similarity calculation unit, used to calculate the similarity between the current user's education experience information and the reference education information; Obtaining the average of the calculated similarities, and determining that the recognition is successful when the average reaches a preset average threshold; The word positioning unit is used to locate the words of the education experience information according to the preset keyword table to obtain static ability keywords.

[0011] As a further solution of the present invention: the dynamic word generation module includes: A file segmentation unit is used to receive the experience information file input by the user, identify the experience information file, and segment it into project files containing time tags; a radius determination unit, configured to obtain a description file of any project file, identify the description file, determine a project difficulty, and determine a project radius according to the project difficulty; A construction unit, configured to obtain capability requirements of a project file, determine display parameters according to the capability requirements, and construct an experience unit of the project file according to the project radius and the display parameters; The first statistical unit is used to count the experience units of the same project file to obtain the experience characteristics; The second statistical unit is used to count all experience features to obtain an experience unit set, select feature units from the experience unit set, obtain the project file corresponding to the feature unit, identify the project file, and obtain dynamic capability keywords; The process of selecting the characteristic unit is as follows: Calculate the difference in display parameters of each pixel point in any two experience units; The parameter differences are cumulatively displayed as unit distances; Calculate the mean unit distance between each experience unit and other experience units, arrange the experience units in increasing order of the mean unit distance, and select a preset number of experience units in sequence as feature units.

[0012] As a further solution of the present invention: the push probability determination module includes: A default value setting unit, used to set the correction coefficient of the static ability keyword to a default value; A correction coefficient determination unit, configured to obtain a time tag of a dynamic capability keyword and determine a correction coefficient based on a radius, display parameters, and time tag of a feature unit of a project file corresponding to the dynamic capability keyword; A matching degree calculation unit, used to calculate the matching degree between each test question in a preset test question library and the static ability keyword and the dynamic ability keyword; An adjustment execution unit, used to adjust the matching degree according to the correction coefficient to determine the comprehensive matching degree of the test questions; A matching degree application unit, configured to determine a probability of pushing a test question based on the comprehensive matching degree; The process of determining the comprehensive matching degree is as follows: Where, For comprehensive matching, is the default value of the correction factor, is the total number of static capability keywords, For the The matching degree of static ability keywords; is the total number of dynamic capability keywords, is the adjustment constant of the dynamic capability keyword, For the The matching degree of static ability keywords, For the Display parameters of static ability keywords, For standard display parameters, For the The time difference between the time tag of a static ability keyword and the current moment, For the The matching degree of static ability keywords.

[0013] Compared with the existing technology, the beneficial effects of the present invention are: the present invention determines keywords based on the user's information file, matches test questions in the question bank based on the keywords, obtains the user's response information, and updates the test question matching process in real time based on the response information, thereby greatly improving the fit between the test questions and the user. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention.

[0015] Figure 1 Flow chart of the personalized test question generation method.

[0016] Figure 2 This is the first sub-flow chart of the personalized test question generation method.

[0017] Figure 3 This is the second sub-flow chart of the personalized test question generation method.

[0018] Figure 4 This is the third sub-flow chart of the personalized test question generation method.

[0019] Figure 5 This is the fourth sub-flow chart of the personalized test question generation method.

[0020] Figure 6 This is a structural block diagram of the personalized test question generation system. DETAILED DESCRIPTION

[0021] In order to make the technical problems, technical solutions and beneficial effects to be solved by the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0022] Figure 1 This is a flowchart of a personalized test question generation method. In an embodiment of the present invention, a personalized test question generation method includes: Step S100: obtaining a user's capability information file, identifying the capability information file, and generating static capability keywords; The technical solution of the present invention is generally applied to the job-seeking field. The user refers to a job seeker. The user's certificate information, retained learning experience and its score (rating) are obtained, which is called a capability information file. The capability information file is identified to obtain keywords used to reflect the user's capabilities, which are called static capability keywords. Most of the capability information file is learning experience, including academic certificates, skill certificates and courses that have passed the assessment.

[0023] Step S200: obtaining a user's experience information file, identifying the experience information file, and generating dynamic capability keywords; The user's experience information file refers to the project file that the user has experienced in the learning and work process. It is used to represent which projects the user has gone through and what results the user has achieved. These are equivalent to experience and change in real time. As time goes by, the experience is constantly updated. The words that reflect the user's abilities are identified in the experience information file and are called dynamic ability keywords.

[0024] Step S300: selecting a question from a preset question bank based on the static ability keywords and the dynamic ability keywords, and simultaneously determining the push probability of the question; Static ability keywords and dynamic ability keywords jointly reflect the user's ability. Test questions are selected from the preset test question bank based on the static ability keywords and dynamic ability keywords as test questions. During the selection process, the matching degree of each test question with the static ability keywords and dynamic ability keywords will be calculated, and the possibility of selecting the test question will be determined based on the matching degree, which is called the push probability.

[0025] Step S400: Pushing test questions based on the push probability, obtaining the user's answer information, generating an evaluation value based on the answer information, and synchronously adjusting the push process; The test questions are pushed based on the push probability. After the test questions are pushed, the reply information input by the user is obtained. The reply information is the user's answer to the test question. The user's ability is evaluated based on the reply information, and the push process is adjusted to finally determine the user's actual ability.

[0026] It is worth mentioning that the dynamic ability keywords all contain time tags. In the process of selecting test questions, the application process of dynamic ability keywords introduces weights based on time tags; the process of determining the weights is: obtaining the time difference between the time tag and the current moment, and determining the weight according to the inverse proportion of the time difference. The larger the time difference, the smaller the weight, which means that the more recent the project experience, the more important it is, and the corresponding weight is greater.

[0027] Figure 2 This is a first sub-flow diagram of the personalized test question generation method. The steps of obtaining the user's ability information file, identifying the ability information file, and generating static ability keywords include: Step S101: Obtaining the education experience information sent by the user, matching similar users based on the education experience information, and querying the education experience information of similar users within a preset time range as reference education information; Step S102: Calculate the similarity between the current user's education experience information and the reference education information; Step S103: obtaining the average of the calculated similarities, and determining that the recognition is successful when the average reaches a preset average threshold; Step S104: locate the words of the education experience information according to the preset keyword table to obtain static ability keywords.

[0028] In an example of the technical solution of the present invention, the generation process of static ability keywords is described. Under normal circumstances, the educational experience information sent by the user is obtained, and the signature verification is performed on it. After the verification is passed, the text recognition algorithm is applied to extract keywords from the educational experience information as static ability keywords; on this basis, the present invention optimizes the verification process, obtains the educational experience information sent by the user, matches similar users based on the educational experience information, calculates the similarity between the current user's educational experience information and the reference educational information, obtains the mean of the calculated similarities, and when the mean reaches a preset mean threshold, it is determined that the recognition is passed; the principle of this process is to calculate the similarity between the educational experience information of the current user and similar users. The advantage of this process is that there is no need to pre-extract standard signature information of different units, and only a sufficient number of user samples are required. As the number of users using the present invention increases, this identification method will become more and more convenient, and it will not have the phenomenon of recognition failure caused by lack of signatures.

[0029] Specifically, there are two ways to match similar users. One is to compare the user's basic information, such as educational background, etc. The other is to directly compare the educational experience information, that is, directly calculate the similarity and select the educational experience information whose similarity reaches a preset similarity threshold as the reference educational experience information. At this time, there is no need to calculate the similarity again and it can be read directly. In the existing technology, educational experience information is generally digital information, such as retained in image format. The similarity calculation process is image similarity, and the image similarity can use the existing similarity calculation scheme with no size limit.

[0030] After the recognition is passed, each educational experience information is traversed according to the preset keyword table. When a word is matched, the position of the matched word is obtained, and the corresponding word is the static ability keyword.

[0031] Figure 3 This is a second sub-flow diagram of the personalized test question generation method. The steps of obtaining the user's experience information file, identifying the experience information file, and generating dynamic ability keywords include: Step S201: receiving an experience information file input by a user, identifying the experience information file, and dividing it into project files containing time tags; Step S202: Obtain a description file of any project file, identify the description file, determine the project difficulty, and determine the project radius according to the project difficulty; Step S203: Acquire capability requirements of the project file, determine display parameters according to the capability requirements, and construct experience units of the project file according to the project radius and the display parameters; Step S204: Count the experience units of the same project file to obtain experience features; Step S205: Count all experience features to obtain an experience unit set, select feature units from the experience unit set, obtain the project file corresponding to the feature unit, identify the project file, and obtain dynamic capability keywords.

[0032] In an example of the technical solution of the present invention, the generation process of dynamic capability keywords is explained, and an experience information file input by a user is received. The experience information file includes multiple projects, and each project contains a time tag. The experience information file is identified and divided into project files containing time tags. It should be noted that the experience information file is generally some text description information, and the recognition process can adopt a conventional text recognition scheme. In fact, there are generally clear separators between different projects. The process of splitting the experience information file into multiple project files is not complicated. In addition, the last moment of the project file is generally selected as the time tag.

[0033] Obtain the description file of any project file, identify the description file, and determine the project difficulty. The project difficulty identification process can adopt the form of table + AI, locate the feature words in the description file, query the difficulty corresponding to the feature words in the preset difficulty table, and then make a comprehensive determination. The difficulty table contains feature words and difficulty items; when an unknown word is detected (a word not included in the common vocabulary), identify the description file based on AI to determine the project difficulty, and determine the project radius based on the project difficulty. The greater the project difficulty, the larger the project radius.

[0034] Acquire the capability requirements of the project file and determine the display parameters based on the capability requirements. The values of the capability requirements are limited. A display parameter is pre-set for each capability requirement. After acquiring the capability requirements of the project file, match the display parameters according to the capability requirements. Construct a circular area based on the project radius and the display parameters, which is called the experience unit of the project file. Count the experience units of the same project file to obtain the experience features, which is equivalent to constructing an experience map. The same user may have multiple project files, each of which corresponds to an experience feature. Count all experience features to obtain an experience unit set. Select feature units from the experience unit set, obtain the project file corresponding to the feature unit, identify the project file, and obtain dynamic capability keywords. It is worth mentioning that the display parameters generally use a single value, such as a grayscale value. Different capability requirements correspond to different grayscale values. Multiple capability requirements correspond to multiple grayscale values. When calculating, they can be superimposed. If the sum exceeds 255 (rare), a range fitting is required to convert the value to 255.

[0035] The process of selecting the characteristic unit is as follows: Calculate the difference in display parameters of each pixel point in any two experience units; The parameter differences are cumulatively displayed as unit distances; Calculate the mean unit distance between each experience unit and other experience units, arrange the experience units in increasing order of the mean unit distance, and select a preset number of experience units in sequence as feature units.

[0036] The process of selecting feature units is not complicated. The difference between each experience unit and other experience units is calculated, represented by the unit distance. The mean of the unit distance is calculated to represent the average difference between each experience unit and other experience units. The experience units are arranged in increasing order according to the average difference, and the experience units are selected from front to back in order as feature units until the number of feature units is large enough.

[0037] Figure 4 This is a third sub-flow diagram of the personalized test question generation method. The steps of selecting test questions from a preset test question library based on static ability keywords and dynamic ability keywords and simultaneously determining the push probability of the test questions include: Step S301: setting the correction coefficient of the static capability keyword to a default value; Step S302: obtaining a time tag of a dynamic capability keyword, and determining a correction coefficient according to the radius, display parameters, and time tag of a feature unit of a project file corresponding to the dynamic capability keyword; Step S303: Calculating the matching degree between each question in the preset question bank and the static ability keyword and the dynamic ability keyword; Step S304: adjusting the matching degree according to the correction coefficient to determine the comprehensive matching degree of the test question; Step S305: Determine the push probability of the test question based on the comprehensive matching degree.

[0038] In an example of the technical solution of the present invention, the application process of keywords is explained, the correction coefficient of the static capability keyword is set to the default value, the time tag of the dynamic capability keyword is obtained, the correction coefficient is determined according to the radius, display parameters and time tag of the feature unit of the project file corresponding to the dynamic capability keyword, and the matching degree of each test question in the preset test question library with the static capability keyword and the dynamic capability keyword is calculated. The matching degree calculation process is the word comparison process. Generally, the keywords are compared with the summary of the test question, and the matching degree is determined based on the number of matches. The more the number, the greater the matching degree. After the matching degree is calculated, the matching degree is adjusted according to the correction coefficient to determine the comprehensive matching degree of the test question. The push probability of the test question is determined according to the comprehensive matching degree. The greater the comprehensive matching degree, the higher the push probability.

[0039] The process of determining the comprehensive matching degree is as follows: Where, For comprehensive matching, is the default value of the correction factor, is the total number of static capability keywords, For the The matching degree of static ability keywords; is the total number of dynamic capability keywords, is the adjustment constant of the dynamic capability keyword, For the The matching degree of static ability keywords, For the Display parameters of static ability keywords, For standard display parameters, For the The time difference between the time tag of a static ability keyword and the current moment, For the The matching degree of static ability keywords.

[0040] The calculation process of comprehensive matching degree is very simple, which is to calculate the product of correction coefficient and matching degree, and then add them together. The calculation process of static ability keywords is relatively simple, and the calculation process of dynamic ability keywords is slightly more complicated, mainly in the correction coefficient. The correction coefficient is proportional to the matching degree, proportional to the difference between the display parameters and the standard display parameters, and inversely proportional to the time difference. and The dimensions used to control static capability keywords and dynamic capability keywords are the same, so that the value ranges belong to the same interval.

[0041] Figure 5 This is a fourth sub-flow diagram of the personalized test question generation method. The steps of pushing test questions based on push probability, obtaining user response information, generating an evaluation value based on the response information, and synchronously adjusting the push process include: Step S401: randomly pushing test questions based on the push probability until the test questions meet a preset condition; the condition is that the test questions of different difficulty levels reach a preset number; Step S402: Obtain the user's response information for each test question, verify the response information based on the reference answer, and generate an evaluation value; Step S403: adjusting the number of test questions of different difficulty levels according to the evaluation value.

[0042] In an example of the technical solution of the present invention, the adjustment process of the push process is to randomly push test questions based on the push probability until the number of test questions of each difficulty meets the preset number, obtain the user's answer information for each test question, verify the answer information based on the reference answer, and generate an evaluation value. The process of generating the evaluation value is the process of comparing the answer information with the reference answer. In the existing intelligent scoring software, it is a basic function and will not be explained here. In layman's terms, the evaluation value is the test question score. The number of test questions of different difficulty levels is adjusted according to the evaluation value. If the evaluation value is large, the number of test questions of the corresponding difficulty is reduced. On the contrary, if the evaluation value is small, the number of test questions of the corresponding difficulty is also reduced. If the evaluation value is close to the preset middle value, the number of test questions of the corresponding difficulty is increased. The function of this process is to make the test questions more questions that the user has mastered but not mastered proficiently, and ultimately determine a more accurate user evaluation situation.

[0043] It is worth mentioning that after the test is completed, each evaluation value needs to be amplified according to the difficulty, and the user's final evaluation result is determined based on the amplified evaluation value.

[0044] As a preferred embodiment of the technical solution of the present invention, the content of the synchronous adjustment push process also includes: Read the evaluation value of each test question; When the evaluation value is less than the preset evaluation threshold, query the dynamic ability keyword corresponding to the test question; Query the project files and experience units corresponding to dynamic capability keywords; The experience unit is updated according to the evaluation value, and the generation process of the feature unit is updated synchronously.

[0045] In an example of the technical solution of the present invention, the adjustment process of the push process is further limited, and the evaluation value of each test question is read. When the evaluation value is small enough, the dynamic capability keyword corresponding to the test question is queried, and the project file and experience unit corresponding to the dynamic capability keyword are queried. The experience unit is updated according to the evaluation value. After the experience unit is updated, the feature unit selection process in step S205 is also updated. The original feature unit may no longer be a feature unit. At this time, the static keyword will also change accordingly, and the test question will also change accordingly. The function of this process is that if the evaluation value of a test question is too low, it will be traced back to the corresponding project file, and the corresponding project file will be used as a reference for generating test questions as much as possible.

[0046] Figure 6 : This is a structural block diagram of a personalized test question generation system. In an embodiment of the present invention, a personalized test question generation system 10 includes: The static word generation module 11 is used to obtain the user's ability information file, identify the ability information file, and generate static ability keywords; The dynamic word generation module 12 is used to obtain the user's experience information file, identify the experience information file, and generate dynamic ability keywords; The push probability determination module 13 is used to select test questions from a preset test question library based on the static ability keywords and the dynamic ability keywords, and simultaneously determine the push probability of the test questions; A negative feedback adjustment module 14 is used to push test questions based on the push probability, obtain the user's answer information, generate an evaluation value based on the answer information, and synchronously adjust the push process; Among them, the dynamic ability keywords all contain time tags. In the process of selecting test questions, the application process of the dynamic ability keywords introduces weights based on time tags.

[0047] Furthermore, the static word generation module 11 includes: A similar user matching unit is used to obtain the educational experience information sent by the user, match similar users based on the educational experience information, and query the educational experience information of similar users within a preset time range as reference educational information; A similarity calculation unit, used to calculate the similarity between the current user's education experience information and the reference education information; Obtaining the average of the calculated similarities, and determining that the recognition is successful when the average reaches a preset average threshold; The word positioning unit is used to locate the words of the education experience information according to the preset keyword table to obtain static ability keywords.

[0048] Specifically, the dynamic word generation module 12 includes: A file segmentation unit is used to receive the experience information file input by the user, identify the experience information file, and segment it into project files containing time tags; a radius determination unit, configured to obtain a description file of any project file, identify the description file, determine a project difficulty, and determine a project radius according to the project difficulty; A construction unit, configured to obtain capability requirements of a project file, determine display parameters according to the capability requirements, and construct an experience unit of the project file according to the project radius and the display parameters; The first statistical unit is used to count the experience units of the same project file to obtain the experience characteristics; The second statistical unit is used to count all experience features to obtain an experience unit set, select feature units from the experience unit set, obtain the project file corresponding to the feature unit, identify the project file, and obtain dynamic capability keywords; The process of selecting the characteristic unit is as follows: Calculate the difference in display parameters of each pixel point in any two experience units; The parameter differences are cumulatively displayed as unit distances; Calculate the mean unit distance between each experience unit and other experience units, arrange the experience units in increasing order of the mean unit distance, and select a preset number of experience units in sequence as feature units.

[0049] Furthermore, the push probability determination module 13 includes: A default value setting unit, used to set the correction coefficient of the static ability keyword to a default value; A correction coefficient determination unit, configured to obtain a time tag of a dynamic capability keyword and determine a correction coefficient based on a radius, display parameters, and time tag of a feature unit of a project file corresponding to the dynamic capability keyword; A matching degree calculation unit, used to calculate the matching degree between each test question in a preset test question library and the static ability keyword and the dynamic ability keyword; An adjustment execution unit, used to adjust the matching degree according to the correction coefficient to determine the comprehensive matching degree of the test questions; A matching degree application unit, configured to determine a probability of pushing a test question based on the comprehensive matching degree; The process of determining the comprehensive matching degree is as follows: Where, For comprehensive matching, is the default value of the correction factor, is the total number of static capability keywords, For the The matching degree of static ability keywords; is the total number of dynamic capability keywords, is the adjustment constant of the dynamic capability keyword, For the The matching degree of static ability keywords, For the Display parameters of static ability keywords, For standard display parameters, For the The time difference between the time tag of a static ability keyword and the current moment, For the The matching degree of static ability keywords.

[0050] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A personalized test question generation method, characterized in that: The method comprises: Obtaining a user's capability information file, identifying the capability information file, and generating static capability keywords; Obtaining a user's experience information file, identifying the experience information file, and generating dynamic capability keywords; Select questions from the preset question bank based on static and dynamic ability keywords, and simultaneously determine the probability of pushing the questions; Push questions based on the push probability, obtain the user's answer information, generate an evaluation value based on the answer information, and synchronously adjust the push process; Among them, the dynamic ability keywords all contain time tags. In the process of selecting test questions, the application process of the dynamic ability keywords introduces weights based on time tags.

2. The personalized test question generation method according to claim 1, characterized in that: The steps of obtaining the user's capability information file, identifying the capability information file, and generating static capability keywords include: Obtain the educational experience information sent by the user, match similar users based on the educational experience information, and query the educational experience information of similar users within a preset time range as reference educational information; Calculate the similarity between the current user's education experience information and the reference education information; Obtaining the average of the calculated similarities, and determining that the recognition is successful when the average reaches a preset average threshold; The educational experience information is word-located according to the preset keyword table to obtain static ability keywords.

3. The personalized test question generation method according to claim 1, characterized in that: The steps of obtaining the user's experience information file, identifying the experience information file, and generating dynamic capability keywords include: Receive the experience information file input by the user, identify the experience information file, and divide it into project files with time tags; Obtain a description file of any project file, identify the description file, determine the project difficulty, and determine the project radius based on the project difficulty; Acquiring capability requirements of a project file, determining display parameters according to the capability requirements, and constructing experience units of the project file according to the project radius and the display parameters; Count the experience units of the same project file to obtain the experience characteristics; Count all experience features to obtain an experience unit set, select feature units from the experience unit set, obtain the project file corresponding to the feature unit, identify the project file, and obtain dynamic capability keywords; The process of selecting the characteristic unit is as follows: Calculate the difference in display parameters of each pixel point in any two experience units; The parameter differences are cumulatively displayed as unit distances; Calculate the mean unit distance between each experience unit and other experience units, arrange the experience units in increasing order of the mean unit distance, and select a preset number of experience units in sequence as feature units.

4. The personalized test question generation method according to claim 3, characterized in that: The step of selecting a question from a preset question bank based on the static ability keywords and the dynamic ability keywords and simultaneously determining the push probability of the question includes: Set the modifiers for static ability keywords to default values; Obtaining a time tag of a dynamic capability keyword, and determining a correction coefficient according to a radius, display parameters, and time tag of a feature unit of a project file corresponding to the dynamic capability keyword; Calculate the matching degree between each question in the preset question bank and the static ability keywords and dynamic ability keywords; Adjust the matching degree according to the correction coefficient to determine the comprehensive matching degree of the test questions; Determining the probability of pushing the test question based on the comprehensive matching degree; The process of determining the comprehensive matching degree is as follows: Where, For comprehensive matching, is the default value of the correction factor, is the total number of static capability keywords, For the The matching degree of static ability keywords; is the total number of dynamic capability keywords, is the adjustment constant of the dynamic capability keyword, For the The matching degree of static ability keywords, For the Display parameters of static ability keywords, For standard display parameters, For the The time difference between the time tag of a static ability keyword and the current moment, For the The matching degree of static ability keywords.

5. The personalized test question generation method according to claim 4, characterized in that: The steps of pushing the test questions based on the push probability, obtaining the user's answer information, generating an evaluation value according to the answer information, and synchronously adjusting the push process include: Randomly push test questions based on the push probability until the test questions meet the preset conditions; the condition is that the test questions of different difficulty levels reach the preset number; Obtain the user's response information for each test question, verify the response information based on the reference answer, and generate an evaluation value; Adjust the number of test questions of different difficulty levels according to the evaluation value.

6. The personalized test question generation method according to claim 5, characterized in that: The synchronous adjustment push process also includes: Read the evaluation value of each test question; When the evaluation value is less than the preset evaluation threshold, query the dynamic ability keyword corresponding to the test question; Query the project files and experience units corresponding to dynamic capability keywords; The experience unit is updated according to the evaluation value, and the generation process of the feature unit is updated synchronously.

7. A personalized test question generation system, characterized in that: The system comprises: A static keyword generation module is used to obtain a user's capability information file, identify the capability information file, and generate static capability keywords; A dynamic word generation module is used to obtain the user's experience information file, identify the experience information file, and generate dynamic ability keywords; A push probability determination module is used to select test questions from a preset test question bank based on static ability keywords and dynamic ability keywords, and simultaneously determine the push probability of the test questions; A negative feedback adjustment module is used to push test questions based on the push probability, obtain the user's response information, generate an evaluation value based on the response information, and synchronously adjust the push process; Among them, the dynamic ability keywords all contain time tags. In the process of selecting test questions, the application process of the dynamic ability keywords introduces weights based on time tags.

8. The personalized test question generation system according to claim 7, characterized in that: The static word generation module includes: A similar user matching unit is used to obtain the educational experience information sent by the user, match similar users based on the educational experience information, and query the educational experience information of similar users within a preset time range as reference educational information; A similarity calculation unit, used to calculate the similarity between the current user's education experience information and the reference education information; Obtaining the average of the calculated similarities, and determining that the recognition is successful when the average reaches a preset average threshold; The word positioning unit is used to locate the words of the education experience information according to the preset keyword table to obtain static ability keywords.

9. The personalized test question generation system according to claim 7, characterized in that: The dynamic word generation module includes: A file segmentation unit is used to receive the experience information file input by the user, identify the experience information file, and segment it into project files containing time tags; a radius determination unit, configured to obtain a description file of any project file, identify the description file, determine a project difficulty, and determine a project radius according to the project difficulty; A construction unit, configured to obtain capability requirements of a project file, determine display parameters according to the capability requirements, and construct an experience unit of the project file according to the project radius and the display parameters; The first statistical unit is used to count the experience units of the same project file to obtain the experience characteristics; The second statistical unit is used to count all experience features to obtain an experience unit set, select feature units from the experience unit set, obtain the project file corresponding to the feature unit, identify the project file, and obtain dynamic capability keywords; The process of selecting the characteristic unit is as follows: Calculate the difference in display parameters of each pixel point in any two experience units; The parameter differences are cumulatively displayed as unit distances; Calculate the mean unit distance between each experience unit and other experience units, arrange the experience units in increasing order of the mean unit distance, and select a preset number of experience units in sequence as feature units.

10. The personalized test question generation system according to claim 9, characterized in that: The push probability determination module includes: A default value setting unit, used to set the correction coefficient of the static ability keyword to a default value; A correction coefficient determination unit, configured to obtain a time tag of a dynamic capability keyword and determine a correction coefficient based on a radius, display parameters, and time tag of a feature unit of a project file corresponding to the dynamic capability keyword; A matching degree calculation unit, used to calculate the matching degree between each test question in a preset test question library and the static ability keyword and the dynamic ability keyword; An adjustment execution unit, used to adjust the matching degree according to the correction coefficient to determine the comprehensive matching degree of the test questions; A matching degree application unit, configured to determine a probability of pushing a test question based on the comprehensive matching degree; The process of determining the comprehensive matching degree is as follows: Where, For comprehensive matching, is the default value of the correction factor. is the total number of static capability keywords, For the The matching degree of static ability keywords; is the total number of dynamic capability keywords, is the adjustment constant of the dynamic capability keyword, For the The matching degree of static ability keywords, For the Display parameters of static ability keywords, For standard display parameters, For the The time difference between the time tag of a static ability keyword and the current moment, For the The matching degree of static ability keywords.

Images