CN117574876A

CN117574876A - Diagnostic report generation method, system and equipment

Info

Publication number: CN117574876A
Application number: CN202410060525.8A
Authority: CN
Inventors: 牟欣; 刘宏涛; 霍晓峰; 郎咸慧
Original assignee: Beijing University of Technology
Current assignee: Beijing University of Technology
Priority date: 2024-01-16
Filing date: 2024-01-16
Publication date: 2024-02-20
Anticipated expiration: 2044-01-16
Also published as: CN117574876B

Abstract

The application discloses a diagnostic report generation method, a system and equipment, which are applied to the technical field of computers, and are used for generating a written text diagnostic report for representing the mastering condition of students on realization attributes based on a first matrix and a cognitive diagnostic model by acquiring the first matrix and the cognitive diagnostic model, wherein the first matrix is used for representing the response condition of written texts of all students in a target student set to items representing advantages and/or disadvantages in the written texts, and the cognitive diagnostic model is a model constructed based on the first matrix, the second matrix and a simplified re-parameterization unified model and is adjusted to be in accordance with preset standards; the second matrix is used for representing the association relation between the project and the realization attribute of the project for representing the requisite thinking ability of the project. The specific quantitative evaluation of the thinking ability in English writing is realized, the whole course can be perfected and adjusted according to the diagnosis report, and students can be guided in a targeted manner.

Description

Diagnostic report generation method, system and equipment

Technical Field

The present disclosure relates to the field of computer technologies, and in particular, to a method, a system, and an apparatus for generating a diagnostic report.

Background

Along with the popularization of international large English capability test, the requirements on the thinking ability in English writing are gradually obvious and deep, the culture of the writing ability of college students is mainly focused on basic aspects such as grammar structures, the culture of the thinking ability is negligent, and the fusion of the classrooms for promoting the thinking ability and English writing is a trend.

However, unlike basic aspects such as grammar and vocabulary, the thinking ability is a potential factor which is difficult to measure and evaluate, in the writing text provided by the student, the correcting person can only find that an error exists in the text or the part which meets the writing requirement, the correcting person cannot know the relation between the thinking ability attribute and the representation in the English writing, the thinking ability of the student cannot be directly measured from the writing text provided by the student, the current thinking ability which is lacking by the student cannot be further determined, and the individualized learning of the thinking ability of the student in the English writing is difficult to be promoted.

Therefore, how to quantitatively evaluate the thinking ability in english writing is a technical problem to be solved by those skilled in the art.

Disclosure of Invention

Based on the above problems, the present application provides a diagnostic report generating method, system and device, which are used for quantitatively evaluating the thinking ability in english writing.

In order to solve the above problems, the technical solution provided in the embodiments of the present application is as follows:

the first aspect of the present application provides a diagnostic report generating method, including:

acquiring a first matrix and a cognitive diagnosis model, wherein the first matrix is used for representing the response condition of the writing text of each student in a target student set to each item, the items are used for representing the advantages and/or the defects in the writing text, and the cognitive diagnosis model is constructed based on the first matrix, the second matrix and a simplified re-parameterized unified model and is adjusted to be in accordance with a preset standard; the second matrix is used for representing the association relation between the project and the realization attribute of the project, and the realization attribute is used for representing the necessary thinking ability when responding to the project;

and generating a written text diagnosis report based on the first matrix and the cognitive diagnosis model, wherein the written text diagnosis report is used for representing the mastering condition of the student on the realization attribute.

Optionally, the method for constructing the cognitive diagnostic model includes:

constructing a preparation model based on the first matrix, the second matrix and the simplified reparameterized unified model;

verifying the second matrix based on dimension estimation and fitness estimation, and adjusting the second matrix based on a verification result until the second matrix meets a first preset standard;

Determining a degree of fit of the preliminary model to the data based on the absolute fit statistics and the relative fit statistics;

and adjusting the preparation model based on the fitting degree and the second matrix until a cognitive diagnosis model meeting a second preset standard is obtained.

Optionally, the written text diagnostic report includes a first sub-diagnostic report, and the generating the written text diagnostic report based on the first matrix and the cognitive diagnostic model includes:

determining a third matrix based on the first matrix, the data mining function and the cognitive diagnosis model, wherein the third matrix is used for representing the overall mastering condition of all students in a target student set on each attribute; the third matrix comprises a column for representing the probability that the attribute is not mastered, two columns for representing the probability that the attribute is mastered, and a row for representing the attribute;

a first sub-diagnostic report is generated based on the third matrix.

Optionally, the written text diagnostic report includes a second sub-diagnostic report, and the generating the written text diagnostic report based on the first matrix and the cognitive diagnostic model includes:

determining an association relation between an attribute mode of an implementation attribute and a target item response probability based on the first matrix, a two-dimensional line drawing function and the cognitive diagnosis model, wherein the attribute mode is a result formed by combining preset parameters based on at least two implementation attributes, and the preset parameters are used for representing whether the implementation attributes are mastered or not;

And generating a second sub-diagnosis report based on the determined association relationship.

Optionally, the written text diagnostic report includes a third sub-diagnostic report, and the generating the written text diagnostic report based on the first matrix and the cognitive diagnostic model includes:

determining the ratio of the number of students corresponding to each attribute mode of the target item to the total number of students corresponding to the student set based on the first matrix, the data mining function and the cognitive diagnostic model;

a third sub-diagnostic report is generated based on the determined ratio.

Optionally, before the obtaining the first matrix and the cognitive diagnostic model, the method further includes:

obtaining correction results corresponding to all students in a target student set, wherein the correction results comprise labeling information corresponding to writing texts;

and generating a first matrix based on the labeling information in each correction result.

Optionally, the labeling information includes an item number, and the generating a first matrix based on the labeling information in each correction result includes:

and integrating and storing project numbers in the correction results of all students in the target student set with the student information to generate a first matrix, wherein the first matrix comprises a first axis used for representing the student information and a second axis used for representing the project numbers, and the first axis and the second axis are transverse and longitudinal axes of the first matrix.

Optionally, the labeling information includes description information for characterizing the content of the project, and after the correction result corresponding to each student in the target student set is obtained, the method further includes:

and combining the selected text corresponding to the labeling information with the descriptive information corresponding to each selected text to generate a correction report corresponding to each student, wherein the correction report comprises the written text and the descriptive information corresponding to all the selected texts in the written text.

A second aspect of the present application provides a diagnostic report generating system comprising:

the system comprises a first acquisition unit, a first matrix and a cognitive diagnosis model, wherein the first matrix is used for representing the response condition of a written text of each student in a target student set to each item, the items are used for representing advantages and/or defects in the written text, and the cognitive diagnosis model is constructed based on the first matrix, the second matrix and a simplified and reparameterized unified model and is adjusted to be in accordance with a preset standard; the second matrix is used for representing the association relation between the project and the realization attribute of the project, and the realization attribute is used for representing the necessary thinking ability when responding to the project;

the first generation unit is used for generating a written text diagnosis report based on the first matrix and the cognitive diagnosis model, and the written text diagnosis report is used for representing the mastering condition of the student on the realization attribute.

A third aspect of the present application provides an electronic device, comprising: the diagnostic report generating method according to any one of the preceding first aspects is implemented by a memory, a processor, and a computer program stored on the memory and executable on the processor, when the processor executes the computer program.

Compared with the prior art, the application has the following beneficial effects:

generating a writing text diagnosis report for representing the mastering condition of a student on realization attributes based on a first matrix and a cognitive diagnosis model by acquiring the first matrix and the cognitive diagnosis model, wherein the first matrix is used for representing the response condition of writing texts of all students in a target student set to items representing advantages and/or disadvantages in the writing texts, and the cognitive diagnosis model is constructed based on the first matrix, a second matrix and a simplified reparameterization unified model and is adjusted to be in accordance with a preset standard; the second matrix is used for representing the association relation between the project and the realization attribute of the project for representing the requisite thinking ability of the project, and the text diagnosis report is written. The method comprises the steps of associating the attribute of the thinking ability in English writing with the item through a second matrix, combining each item representing the advantages and disadvantages in the writing text of the student with the thinking ability required to be provided for responding to the item in a matrix form, and analyzing through the item and the cognitive diagnosis model reflected in the text, so that the specific quantitative evaluation of the thinking ability in English writing is realized, and the mastering condition of the student on the realization attribute is obtained. The teacher can perfect and adjust the course according to the diagnosis report, and the teacher can instruct students in a targeted way.

Drawings

In order to more clearly illustrate the present embodiments or the technical solutions in the prior art, the drawings that are required for the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

FIG. 1 is a flowchart of a diagnostic report generating method according to an embodiment of the present application;

FIG. 2 is a schematic diagram of an correction result provided in an embodiment of the present application;

FIG. 3 is a schematic diagram of an correction result provided in an embodiment of the present application;

fig. 4 is a schematic diagram of a first matrix provided in an embodiment of the present application;

FIG. 5 is a schematic diagram of attribute mode and target item response probability provided in an embodiment of the present application;

fig. 6 is a structural diagram of a diagnostic report generating system according to an embodiment of the present application.

Detailed Description

In order to make the present application solution better understood by those skilled in the art, the following description will clearly and completely describe the technical solution in the embodiments of the present application with reference to the accompanying drawings in the embodiments of the present application, and it is apparent that the described embodiments are only some embodiments of the present application, not all embodiments. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present disclosure, are within the scope of the present disclosure.

In order to facilitate understanding of the technical solutions provided in the embodiments of the present application, technical terms related to the embodiments of the present application will be described first.

First, english writing thinking ability: for a long time, china focuses on cultivating English ability of students, wherein English writing ability is a very important ring, for college students, basic grammar ability is mastered through English ability training and cultivation for decades, but thinking ability in English writing is always ignored, and the requirements of global large English ability tests (such as yasi, tuofu, GRE and the like) clearly suggest that examinees should show better thinking ability in English writing, so that thinking ability of measuring, cultivating and training Chinese college students in English writing becomes a very important part in English teaching of colleges and universities.

Second, cognitive Diagnostic Model (CDM): cognitive diagnostics are psychological and educational measures characterized by a combination of model-based measures and formative assessment, and test results of cognitive diagnostics can be predictive of the interpretability of a particular field. Cognitive diagnostic assessment aims at measuring a student's specific knowledge structure and processing skills in order to provide information about their cognitive advantages and disadvantages. The cognitive psychologist and the cognitive psychometrist combine the research result and the research paradigm of the cognitive psychology on the internal mechanism of the human cognitive processing process, and creatively develop a cognitive diagnosis model with a cognitive diagnosis function. A cognitive diagnostic model is a measurement model that contains discrete, potential attribute variables associated with a domain, which refers to the grasp of binary skills or knowledge, which represents the individual basic features of the domain covered by a test item. The cognitive diagnostic model may estimate a student's specific capacity profile by observing answers to certain attribute variables of the subject.

In order to facilitate understanding of the technical solutions provided in the embodiments of the present application, the following description will first explain the background technology related to the embodiments of the present application.

Firstly, as described above, the culture of writing ability of college students is mainly focused on basic aspects such as grammar structures, the culture of writing ability is neglected, but with the popularization of international large-scale English ability test, the requirement of writing ability in English writing is gradually obvious and deep, and the fusion of writing ability and English writing is promoted. However, unlike basic aspects such as grammar and vocabulary, the thinking ability is a potential factor which is difficult to measure and evaluate, and on one hand, a teacher cannot determine how to teach the thinking ability in writing, and on the other hand, the teacher cannot measure the thinking ability which is exhibited in English writing of students. Second, how to define the attributes and characterizations of the thinking ability in English writing is a problem. There are some studies on definition of cognitive diagnostic properties of grammar structures in english writing, and at the same time there is some accumulation of studies on the mind-distinguishing ability in the field of physics, but there is a lack of study on fusion of both.

To solve this problem, the embodiment of the application provides a diagnostic report generation method, a diagnostic report generation system and diagnostic report generation equipment. First, a cognitive diagnostic model that can evaluate potential features, quantifying feature attributes, presents its advantages. The method is characterized in that the thinking ability attribute and the characterization are defined, the Q matrix is generated, and the proper cognitive diagnosis model is selected, so that the quantitative analysis of students on the thinking ability attribute can be obtained, the teacher is facilitated to design and realize classroom contents, meanwhile, the students are helped to know the weak parts of the students, and the individualized learning of the students on the thinking ability in English writing is promoted. The application provides an initial Q matrix (0/1 matrix analysis and resolution capability attribute and a representation corresponding relation in English writing); then, the initial Q matrix is discussed, analyzed and corrected to obtain a Q matrix (namely a second matrix) suitable for college students to measure the thinking-distinguishing capability in English writing; and finally, carrying out relevant verification on the Q matrix in the cognitive diagnosis model, wherein the relevant verification comprises dimension verification, relative fitting comparison and the like, and carrying out mathematical quantitative fine adjustment on the Q matrix so that the Q matrix is more suitable for the cognitive diagnosis model.

Summarizing, i.e. no matrix for the legibility in english writing is currently available, the scheme determines the Q matrix (second matrix) by quantization analysis; at present, no cognition diagnosis model is selected for the evaluation of the thinking ability in English writing, and the scheme verifies that RRUM can be used as the cognition diagnosis model of the subject through data acquisition, data conversion analysis, experiment implementation and the like; at present, the thinking ability in English writing cannot be quantitatively evaluated, and the method can obtain a specific quantitative analysis evaluation report of the thinking ability of students in English writing by defining a Q matrix, selecting a proper cognitive diagnosis model and calculating posterior distribution of all aspects.

It should be noted that, in the embodiments of the diagnostic report generating method, system and device provided in the present application, the execution body of the diagnostic report generating method may not be limited, and for example, the diagnostic report generating method in the embodiments of the present application may be applied to a data processing device such as a terminal device or a server. The terminal device may be an electronic device such as a smart phone, a computer, a personal digital assistant (Personal Digital Assistant, PDA), a tablet computer, etc. The server may be an independent server, a cloud server, or a cluster server composed of a plurality of servers.

For the purpose of making the objects, technical solutions and advantages of the embodiments of the present application more clear, the technical solutions of the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is apparent that the described embodiments are only some embodiments of the present application, but not all embodiments. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present disclosure, are within the scope of the present disclosure.

A diagnostic report generating method provided in the present application is described below by way of one embodiment. Referring to fig. 1, the flowchart of a diagnostic report generating method provided in the embodiment of the present application is provided, where an execution subject of the method flow is a server, and further, the subject may be a report generating system in the server, where the method includes:

s101, acquiring a first matrix and a cognitive diagnosis model.

The first matrix is used to characterize the response of the written text of each student in the target student set to each item used to characterize the advantages and/or disadvantages in the written text. In conventional cognitive diagnosis, a test paper, such as 10 questions, is generally arranged, then 10 items are arranged, and the answer pair is the response, but in text writing, the 10 questions are equivalent to being embodied in text content. The teacher is arranged to write a composition to the student, and according to the text given by the student, the teacher is used as a corrector to correct the composition text provided by the student, in the correction process, whether ten items are reflected in the composition text or not is corrected, for example, a student writes a trace "I argue coat …" in the text, the teacher considers that the part of the text shows a "clear main idea of the author", namely, a certain preset item is responded, and the process is called response.

The cognitive diagnosis model is constructed based on the first matrix, the second matrix and the simplified and reparameterized unified model, and is adjusted to a model meeting a preset standard; the second matrix is used for representing the association relation between the project and the realization attribute of the project, and the realization attribute is used for representing the necessary thinking ability when responding to the project.

In one possible implementation manner, before the first matrix and the cognitive diagnostic model are obtained, the method further includes the following steps A1-A2:

and A1, acquiring correction results corresponding to all students in a target student set.

And the correction result comprises annotation information corresponding to the written text. The labeling information is obtained by labeling a written text provided by a student based on a preset text item, and in an actual application scene, the labeling information can comprise an item number matched with the item content and/or descriptive information used for representing the item content.

When the labeling information is presented in the form of item numbers, reference may be made to fig. 2, fig. 2 is a schematic diagram of the correction result provided in the embodiment of the present application, where the left side of the picture is a part of the written text provided by a student in the student set, and the numbers and letters from top to bottom on the right side of the picture are "3", "1", "18", "7", "6", "a", "5", "a" respectively, that is, the item numbers in the labeling information. The "corrector 1" is a user name of the corrector, and can be deleted in an actual application scene.

When the annotation information is presented in the form of description information, reference may be made to fig. 3, where fig. 3 is a schematic diagram of the correction result provided in the embodiment of the present application, where the left side of the picture is a part of the writing text provided by a student in the student set, and the text from top to bottom on the right side of the picture is the description information in the annotation information. The "corrector 1" is a user name of the corrector, and can be deleted in an actual application scene.

It should be noted that, the correspondence between the item numbers and the description information may be shown in the following table 1, where table 1 is a correspondence table between the item numbers and the description information, that is, in fig. 2 and 3, the first annotation "3" in fig. 2 and the first annotation "clearly indicates the main point of the author" in fig. 3 represent the same meaning, and the annotation number 3 corresponds to the description information, and in an actual application scenario, for convenience of identification, the item numbers and the description information may be noted at the annotation information, that is, the annotation may be presented as "3: the main point of view of the author is clearly presented, wherein the connection symbol between the project number and the descriptive information and the arrangement sequence between the project number and the descriptive information can be adaptively adjusted according to the actual requirements.

In an actual application scene, word can be used for carrying out corresponding correction on English writing provided by students, and annotating text representations presented by the students: if English writing of students shows that "introduction is suitable for readers, attractive and innovative" at a place, annotating "1" at the place.

TABLE 1 correspondence table between item numbers and description information

And A2, generating a first matrix based on the labeling information in each correction result.

In the application process, the Python codes can be used for summarizing the numbers of the students-projects, and a summary matrix of the students-projects, namely a first matrix, is generated. In one possible implementation, invalid data may be deleted using Python code, where invalid data refers to all zero data that a student does not respond to all characterizations.

Summarizing the project conditions of all students in the student set into a matrix, wherein the rows of the matrix represent the students, and the columns of the matrix represent the projects, namely, project numbers in the annotation information above, such as 1 indicates that the introduction in the representation 1 is suitable for readers, attractive and innovative); assuming a test evaluatesItems, by->A matrix is formed by students Specify the association between the item and the student, M _n，j Is->Line->The elements on the column represent +.>Individual students are about the project>If the student responds to the item (the response means that the text representation corresponding to the item is embodied in the written text provided by the student, namely, the result after the teacher is corrected), the element is 1, otherwise, the element is 0. FIG. 4 is a schematic diagram of a first matrix provided in an embodiment of the present application, showing Python generationA portion of the matrix. The Python codes are utilized to save a large amount of time for summarizing and organizing data, and the working difficulty and strength of teachers are reduced. In a practical application scenario, the rows and columns may be swapped.

In one possible implementation manner, the method for constructing the cognitive diagnosis model may include:

and A1, constructing a preparation model based on the first matrix, the second matrix and the simplified reparameterized unified model.

The second matrix is used for representing the association relation between the project and the realization attribute of the project, and the realization attribute is used for representing the necessary thinking ability when responding to the project.

In an actual application scene, firstly, generating an initial Q matrix based on an item, and inputting the Q matrix into a separate table to form a Q matrix table, wherein the Q matrix is shown in a table 2, and the table 2 is a corresponding relation table between the item and the realization attribute of the item, namely a second matrix, wherein a row of the matrix represents the item and is listed as an attribute representing the thinking ability, and the matrix reflects the reaction relation between the English text item and the thinking ability; suppose that the current assessment includes Item(s)>Attributes, a matrixSpecify an association between an item and an attribute, where Q _j，a Is->Line->The elements on the column represent +.>Individual items are->If a change in property is required in response to the item, then the element is 1, otherwise the element is 0. This matrix is the key to the cognitive diagnostic model. For example, item 5 in Table 2, a student may only answer item 5 if he has two implementation properties, A1 and A3.

TABLE 2 correspondence table between items and implementation properties of the items

The step is that the R language is used for carrying out cognitive diagnosis analysis on the data, and the GDINA sub-package (Generalized Deterministic Inputs, noise and Gate) is loaded in the R language firstly, and the generalized deterministic input, noise and Gate are inputted; loading the first matrix and the second matrix obtained in the above step by using a GDINA function, and assigning a cognitive diagnostic model of "RRUM": model=gdina (M, Q, model= 'RRUM');

the function builds a cognitive diagnostic model based on RRUM (Reduced Reparameterized Unified Model, simplified re-parameterized unified model) through the student-project reaction matrix M (i.e., the first matrix) and the initial Q matrix.

And A2, verifying the second matrix based on dimension estimation and fitness estimation, and adjusting the second matrix based on a verification result until the second matrix meets a first preset standard.

The current Q matrix is an initially generated Q matrix (i.e., a second matrix before verification), and a mathematical statistical method is required to quantitatively evaluate the Q matrix (i.e., the second matrix before verification) to form a final Q matrix (i.e., the second matrix after verification). In an actual application scene, the pre-set model is constructed based on a second matrix before adjustment, and after the second matrix is adjusted, the preparation model can be adjusted according to the adjusted second matrix, and then a related adjustment process of the subsequent fitting degree is performed.

First, dimension assessment provides valuable information for understanding the structure of the initial Q matrix. The dimensions of the Q matrix are estimated using the Qval function in the GDINA packet: dimension = gdina. Qval (model); wherein, GDINA is the GDINA package in the R software used above, qval is Q matrix Validation (Q matrix verification), and model is the model built by GDINA; the return value dimension of the function is a value for performing dimension estimation on the Q matrix; the value should be the same as the attribute value of the Q matrix. For example, in the present evaluation method, the dimension estimated by using the function is 5, and similarly, the number of attributes is 5, which accords with the verification of the dimension estimation.

Then, several model fitting indexes of different Q-matrices established by the discrete factor loading method (DFL) and the hell method were compared with a specified number of attribute fitting models. The Alcak Information Criterion (AIC) and Bayesian Information Criterion (BIC) are used to evaluate the fitness, which can be specifically evaluated using the AIC function and BIC function in the GDINA package: aic_result=gdina. Bic_result=gdina.bic (model); the model is the model established by the GDINA, the AIC_result shows the AIC value calculated by the model when the attribute takes different values, and the BIC_result shows the BIC value calculated by the model when the attribute takes different values; the smaller the AIC and BIC values, the more appropriate the Q matrix is; that is, in the result, it should be observed that the attribute value corresponding to when AIC and BIC are minimum should be equal to the attribute value of the Q matrix itself; for example, in the present evaluation method, the attribute value of the Q matrix is 5, and it is also observed that AIC and BIC are minimized when the value is equal to 5.

And A3, determining the fitting degree of the preparation model and the data based on absolute fitting statistics and relative fitting statistics.

And step A4, adjusting the preparation model based on the fitting degree until a cognitive diagnosis model meeting a second preset standard is obtained.

Two mathematical statistical methods are used to evaluate the fitness of the model to the data: absolute fit statistics and relative fit statistics. The former examines the fitting degree of the model and the data in absolute terms, while the latter uses a comparison view to select the best model from among many models.

The statistics compare the univariate and bivariate distributions of observations and model predictions. Since the statistics fit the distribution, a hypothesis test can be performed to evaluate whether the model fits the data. However, given that the test is affected by the sample size, large samples may capture negligible differences between the model and the data. To address this problem, the effect size can be measured using the root mean square approximation error (RMSEA 2) and the normalized root mean square residual (SRMSR). For RMSEA2 and SRMSR, a smaller number indicates a higher absolute fit of the model data. Simulation studies showed that RMSEA2<0.03 indicates excellent fit, 0.03< RMSEA2<0.045 indicates good fit, and RMSEA2<0.045 indicates poor fit. SRMSR <0.05 indicates a good model fit. The specific usage functions are: the values are obtained by GDINA. Model, where GDINA is the R package, model is a fit estimate to the model, and model is the model built from the GDINA package. In the present evaluation method, using this function, the result at rmsea2= 0.0265 and srmsr=0.0312 was obtained.

The relative fit statistics using different indices illustrate the best model for the particular data and Q matrix. A series of models were applied to the data and AIC, BIC and observed log-Likelihood Ratio Test (LRT) were calculated to verify whether RRUM (the above-used cognitive diagnostic model) is the best model. Candidate models include G-DINA, DINA, ACDM and RRUM. The model for obtaining the minimum log-likelihood ratio and AIC index shows that the suitability is good. The specific usage functions are: modelcomp (model) GDINA. Modelcomp (model) is a comparative estimate of the various models, where GDINA is the R package described above, and model is the model built by the GDINA package described above. In the present evaluation method, it is observed that the minimum AIC, BIC, etc. parameters are obtained when the model is RRUM, using this function.

In the actual application scenario, the sequence of the two verification steps can be adjusted according to the actual requirement, or can be performed simultaneously. The first preset standard and the second preset standard can be set according to actual requirements.

There are 60 existing cognitive diagnostic models, and how to select a suitable cognitive diagnostic model for quantification of the thinking ability in English writing is a problem. First, a cognitive diagnostic model is classified: the method belongs to a potential classification model after analysis and discussion, wherein the model presumes that skills are discontinuous, a potential skill space consists of a plurality of binary variables to obtain a plurality of cognitive states, and students can be divided into the cognitive states so as to distinguish the tested cognitive structures. The competence model is to solve the problem of assuming skill mastery is binary or multiple categories, most commonly 0, 1 scoring mode. It is also determined how different cognitive attributes are related, whether directly or indirectly, or not. The project structure model also considers whether the interaction between cognitive attributes is compensatory or uncompensated. The G-DINA model distinguishes a population of subjects who do not have all the attributes measured by the problem in full grasp, and no longer assumes that all these subjects have the same answer pair probability. The RRUM model is simplified on the G-DINA model and assumes that there is no compensation between attributes. In combination with this solution, students must master all attributes to answer the corresponding item. Thus, the model should be uncompensated and have no measurement of residual capacity. The method uses RRUM as a cognitive diagnostic model; meanwhile, absolute fit analysis and relative fit analysis are used for carrying out quantitative analysis on the mathematical level on the model, and the use of the RRUM model is determined.

S102, generating a written text diagnosis report based on the first matrix and the cognitive diagnosis model.

The written text diagnostic report is used for representing the mastering condition of the student on the realization attribute.

In one possible implementation, the written text diagnostic report includes a first sub-diagnostic report, and the generating a written text diagnostic report based on the first matrix and the cognitive diagnostic model includes the steps of:

a third matrix is determined based on the first matrix, a data mining function, and a cognitive diagnostic model, wherein a first sub-diagnostic report is generated based on the third matrix.

After verifying the Q matrix and the model, carrying out statistical analysis on the attribute mastering conditions of the whole students in the step, and specifically using functions: GDINA: exacts (model, what= 'prevalce'), which is a data mining function, can analyze an input model and a designated variable, and the model is the model established above, and prevalce refers to the mastering condition of each attribute;

i.e. the third matrix of outputs isThe third matrix is used for representing the overall mastering condition of all students in the target student set on each attribute; the third matrix includes a column that characterizes the probability that the attribute is not known and two columns that characterize the probability that the attribute is known, and a row for characterizing the attribute. The >The row indicates->The mastering condition of the attribute is named as 'Level 0', corresponding to the unoccupied probability of the overall student for the attribute, and the 2 nd column is named as 'Level 1', corresponding to the mastered probability of the overall student for the attribute, and the sum of the two values of each row is 1; when->Level0 of row>And when Level1, the current student has poor grasp of the attribute.

In an actual application scenario, the first sub-diagnostic report may be generated according to values corresponding to at least one row in the third matrix. In one possible implementation, the corresponding first sub-diagnostic reports may be generated for several rows in the third matrix according to actual requirements.

In one possible implementation, the written text diagnostic report includes a second sub-diagnostic report, and the generating the written text diagnostic report based on the first matrix and the cognitive diagnostic model includes:

and determining an association relation between an attribute mode for realizing the attribute and the target item response probability based on the first matrix, the two-dimensional line drawing function and the cognitive diagnosis model, and generating a second sub-diagnosis report based on the determined association relation.

The attribute mode is a result formed by combining preset parameters based on at least two implementation attributes, wherein the preset parameters are used for representing whether the implementation attributes are mastered or not. In an actual application scenario, the preset parameters can be set according to actual requirements, for example, the preset parameters are set to 0 and 1, 1 is used for representing that the student has mastered the realization attribute, and 0 is used for representing that the student has not mastered the realization attribute.

Namely, specific diagnosis analysis is carried out on the project, and the relation of the grasping modes of different attributes to project responses is studied, so that specific relation between the project and attribute grasping can be obtained. The specific usage functions are: GDINA: plot (model, item=j), where model is the model generated above, item refers to the item, representing the current function study as the item. The function outputs the corresponding relation between the mastering condition of the attribute and the item; fig. 5 is a schematic diagram of an attribute mode and a response probability of a target item provided in an embodiment of the present application, where the graph is an analysis of response probability of an item 12, and the horizontal axis is four attribute modes, and the vertical axis is response probability of the target item, that is, when item=12, the output content of the function is shown; in table 1, the answer pair of item 12 needs two realization attributes of A4 and A5, and based on the response probability of the target item corresponding to the four combinations in fig. 4, the attribute mode of the realization attribute in the lower graph is 00, and the combination is used for representing the situation that A4 is not mastered and A5 is not mastered, 10 means that A4 is mastered and A5 is not mastered; 01 refers to mastering A5 without mastering A4; 11 means grasping both A4 and A5. When both attributes are grasped together, that is, when the attribute mode of the realization attribute corresponding to the item 12 is 11, the probability of responding to the item 12 is highest, and when A4 is grasped without A5, that is, when the attribute mode of the realization attribute corresponding to the item 12 is 10, it is found that there is a certain probability of responding to the item 12, and the explanation of the item 12 is more dependent on the A4 attribute.

In an actual application scenario, the second sub-diagnosis report may be generated according to the item response probabilities corresponding to the four attribute modes, that is, presented in the form of "attribute mode-response probability". In one possible implementation manner, the response probability corresponding to each attribute mode output by the model can be further analyzed to obtain which implementation attribute the current target item depends on. The target item may be one or more items selected by the modifier, or may be all preset items, for example, at least one of 22 items in table 1 may be 22 at most.

In one possible implementation, the written text diagnostic report includes a third sub-diagnostic report, and the generating the written text diagnostic report based on the first matrix and the cognitive diagnostic model includes:

determining the ratio of the number of students corresponding to each attribute mode of the target item to the total number of students corresponding to the student set based on the first matrix, the data mining function and the cognitive diagnostic model; a third sub-diagnostic report is generated based on the determined ratio.

And (3) performing potential attribute mode analysis: analyzing the overall mastering situation of the attribute mode, the mastering situation of different attribute combination modes can be seen, such as attribute mode 10010, which represents mastering attributes 1 and 4 and not mastering attributes 2,3 and 5; analyzing the attribute mode to further obtain the mastering condition of the student on the attribute; the specific usage functions are: GDINA: the model is generated as described above, and the model is a posterior distribution probability. From the following components The individual attributes can be obtained->A personal attribute mode; for example, a=2, from 4 attribute modes 00, 01, 10, 11; i.e., neither master, master the first not master the second, master the second not master the first, both master; the function will output +.>The posterior distribution probability corresponding to the attribute mode is specifically, for example, 2 attributes are included, the posterior distribution probability corresponding to 00 refers to that the number of students which are not mastered by both attributes accounts for the total number of students in the student set, the posterior distribution probability of 01, 10 and 11 is the same, and the sum of the 4 probabilities is equal to 1, namely, the overall crowd, so that the distribution situation of the attribute mode in the student set is obtained.

In an actual application scenario, a ratio of the number of students corresponding to each attribute mode to the total number of students corresponding to the student set may be obtained, and a third sub-diagnosis report may be generated and presented in a corresponding manner of "attribute mode-ratio". Further, the calculated ratios may be ranked so that the teacher may more significantly notice the largest student occupancy for that attribute mode.

The three sub-diagnostic reports correspond to different diagnostic directions, and in one possible implementation, the three sub-diagnostic reports may be integrated to obtain an overall diagnostic report, and the overall diagnostic report may include all contents of the three sub-diagnostic reports, that is, may be integrated into an overall diagnostic report on the basis of not changing the contents of each sub-diagnostic report. A new overall diagnostic report may also be generated based on the contents of the three sub-diagnostic reports.

In one possible implementation manner, after the obtaining the correction result corresponding to each student in the target student set, the method further includes:

The annotation information includes descriptive information for characterizing the content of the item. I.e. generating a correction report provided to the student, wherein the selected text is the text sentence corresponding to each annotation in fig. 3, which represents the advantages and/or disadvantages in the written text. The student can obtain a correction report of the descriptive information added with the corresponding annotation on the basis of the self-provided writing text, so that the student can know the related problems of the self-provided writing text.

In summary, the method provided by the application firstly defines and measures the thinking ability attribute in writing by referring to the existing literature, extracts the characterization of the thinking ability in English writing, creatively provides the Q matrix necessary in cognitive diagnosis, and then uses conversion software written by Python to carry out statistical conversion on the corrected writing sample. Firstly, checking a Q matrix based on a cognitive diagnosis model G-DINA realized by using an R language, wherein the checking comprises dimension checking, relative fitting comparison and the like; the model is then corrected using an absolute fit analysis and a relative fit analysis, ultimately yielding an overall, individual-specific quantitative assessment of the sample's ability to identify. By the method, quantitative evaluation analysis of the thinking ability in English writing is effectively realized, and verification is carried out in an experimental sample.

The method is used for evaluating the thinking ability of students in colleges and universities in English writing, and the method realizes the specific quantitative evaluation of the thinking ability in English writing for the first time and generates individual and integral thinking ability evaluation reports; the thinking ability evaluation report of the whole student is beneficial to the teacher to accurately evaluate the thinking ability of the student in English writing, so that courses are perfected pertinently, the method can be used for effectively improving and enhancing the teaching quality of English writing courses, and the thinking ability evaluation report of the individual student can pertinently provide important aspects of later training and cultivation for the student. The method is simple to use, and the teacher can obtain the overall and personal thinking ability assessment report only after correcting the student writing. The teacher can perfect and adjust the course overall according to the overall evaluation report, and the student can be guided in a targeted manner for the personal evaluation report. The development of English writing class of students in colleges and universities in China is promoted, the teaching of knowledge such as basic grammar structure and the like can be improved to training of thinking ability, and the requirements of the current international English level assessment are met, so that the English level assessment method has a wide application prospect.

The foregoing is a specific implementation manner of the diagnostic report generating method provided in the embodiments of the present application, and based on this, the present application further provides a corresponding system for generating a diagnostic report. The system provided in the embodiments of the present application will be described from the viewpoint of functional modularization. Fig. 6 is a structural diagram of a diagnostic report generating system according to an embodiment of the present application.

The system comprises:

a first obtaining unit 110, configured to obtain a first matrix and a cognitive diagnostic model, where the first matrix is used to characterize a response situation of a written text of each student in a target student set to each item, the item is used to characterize advantages and/or disadvantages in the written text, and the cognitive diagnostic model is a model constructed based on the first matrix, the second matrix and a simplified reparameterized unified model, and is adjusted to meet a preset standard; the second matrix is used for representing the association relation between the project and the realization attribute of the project, and the realization attribute is used for representing the necessary thinking ability when responding to the project;

a first generating unit 111, configured to generate a written text diagnosis report based on the first matrix and the cognitive diagnosis model, where the written text diagnosis report is used to characterize the mastering situation of the implementation attribute by the student.

Optionally, the first generating unit is specifically configured to:

determining a third matrix based on the first matrix, the data mining function and the cognitive diagnosis model, wherein the third matrix is used for representing the overall mastering condition of all students in a target student set on each attribute; the third matrix comprises a column for representing the probability that the attribute is not mastered, two columns for representing the probability that the attribute is mastered, and a row for representing the attribute; a first sub-diagnostic report is generated based on the third matrix.

Optionally, the first generating unit is specifically configured to:

a third sub-diagnostic report is generated based on the determined ratio.

Optionally, the system further comprises:

the second acquisition unit is used for acquiring correction results corresponding to all students in the target student set, wherein the correction results comprise annotation information corresponding to the writing text;

and the second generation unit is used for generating a first matrix based on the labeling information in each correction result.

Optionally, the second generating unit is specifically configured to:

Optionally, the system further comprises:

and the third generation unit is used for combining the selected text corresponding to the marking information with the description information corresponding to each selected text to generate a correction report corresponding to each student, wherein the correction report comprises the writing text and the description information corresponding to all the selected texts in the writing text.

The embodiment of the application also provides corresponding equipment and a computer storage medium, which are used for realizing the scheme of the diagnostic report generation method.

The device comprises a memory for storing instructions or code and a processor for executing the instructions or code to cause the device to perform the diagnostic report generation method of any of the embodiments of the present application.

The computer storage medium has code stored therein, and when the code is executed, a device executing the code implements the diagnostic report generating method described in any embodiment of the present application.

It should be noted that, in the present description, each embodiment is described in a progressive manner, and each embodiment is mainly described in a different manner from other embodiments, and identical and similar parts between the embodiments are all enough to refer to each other. For the system or device disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple, and the relevant points refer to the description of the method section.

It should be understood that in this application, "at least one" means one or more, and "a plurality" means two or more. "and/or" for describing the association relationship of the association object, the representation may have three relationships, for example, "a and/or B" may represent: only a, only B and both a and B are present, wherein a, B may be singular or plural. The character "/" generally indicates that the context-dependent object is an "or" relationship. "at least one of" or the like means any combination of these items, including any combination of single item(s) or plural items(s). For example, at least one (one) of a, b or c may represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", wherein a, b, c may be single or plural.

It is further noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.

The steps of a method or algorithm described in connection with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by a processor, or in a combination of the two. The software modules may be disposed in Random Access Memory (RAM), memory, read Only Memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A diagnostic report generation method, comprising:

2. The method of claim 1, wherein the method of constructing the cognitive diagnostic model comprises:

3. The method of claim 1, wherein the written text diagnostic report comprises a first sub-diagnostic report, the generating a written text diagnostic report based on the first matrix and the cognitive diagnostic model comprising:

A first sub-diagnostic report is generated based on the third matrix.

4. The method of claim 1, wherein the written text diagnostic report comprises a second sub-diagnostic report, the generating a written text diagnostic report based on the first matrix and the cognitive diagnostic model comprising:

5. The method of claim 1, wherein the written text diagnostic report includes a third sub-diagnostic report, the generating a written text diagnostic report based on the first matrix and the cognitive diagnostic model comprising:

A third sub-diagnostic report is generated based on the determined ratio.

6. The method of claim 1, wherein prior to the obtaining the first matrix and the cognitive diagnostic model, further comprising:

7. The method of claim 6, wherein the annotation information comprises an item number, and wherein generating the first matrix based on the annotation information in each of the modification results comprises:

8. The method according to claim 6, wherein the labeling information includes description information for characterizing the content of the item, and after the obtaining the correction result corresponding to each student in the target student set, the method further includes:

9. A diagnostic report generating system, the system comprising:

10. An electronic device, comprising: a memory, a processor, and a computer program stored on the memory and executable on the processor, the processor implementing the diagnostic report generation method of any one of claims 1-8 when the computer program is executed.