CN115797122B

CN115797122B - Operation analysis method and device

Info

Publication number: CN115797122B
Application number: CN202211702819.3A
Authority: CN
Inventors: 邵雅清; 李扬; 郑剑飞; 程崇臻; 陈捷; 罗代势; 刘志民; 连一凡
Original assignee: Beijing Hex Technology Co ltd
Current assignee: Beijing Hex Technology Co ltd
Priority date: 2022-12-29
Filing date: 2022-12-29
Publication date: 2024-02-20
Anticipated expiration: 2042-12-29
Also published as: CN115797122A

Abstract

The application provides a job analysis method and a job analysis device, and belongs to the technical field of teaching auxiliary management. The method comprises the steps of obtaining historical homework information of a target student and current homework response information of the target student; and then comparing the historical job information with the job response information to determine whether the job response information is analyzable job response information according to the comparison result of the historical job information and the job response information. If yes, the homework response information of the target students is input into a preset comprehensive evaluation model so as to sequentially match a plurality of analysis evaluation texts of the target students in a preset evaluation dimension set. Wherein evaluating the set of dimensions comprises: student work accuracy dimension, work stability dimension, work time efficiency management dimension, class work accuracy dimension. And generating an operation analysis report of corresponding statistical period according to each analysis and evaluation text and the evaluation word set corresponding to the target student. The set of evaluation mood words includes mood words for matching the expression job analysis report.

Description

Operation analysis method and device

Technical Field

The application relates to the technical field of teaching auxiliary management, in particular to a job analysis method and a device.

Background

With the progress of educational informatization, the trace collection of student work data has now been achieved. In the context of the accompanying electronic collection of student work data, comprehensive analysis feedback about the procedural work level of students is relatively lacking. The learning effect of the student in the learning process can be effectively measured through the procedural homework analysis report, and the student is helped to know own advantages and defects in time, so that the student can exercise accurately in pertinence, and the student performance is improved while the academic burden is not increased.

The applicant found that the statistical index of the homework report is limited to statistics of the homework accuracy of students only when the procedural homework report is displayed, the comparison and analysis angles are single in many cases, and when the homework report is displayed to the students, the students are difficult to resonate the homework report. It is also difficult for students to actively face the homework report results and to actually correct the homework results.

Disclosure of Invention

The embodiment of the application provides an homework analysis method and device, which are used for solving the technical problems that the evaluation of the current student procedural homework report is not objective, students cannot resonate the homework report, knowledge holes of the students cannot be effectively reflected through the homework report, and the learning performance of the students is improved.

In one aspect, an embodiment of the present application provides a job analysis method, including:

and acquiring the historical homework information and the current homework response information of the target students. The history job information comprises the history job answering time and the history job short text of each question type. The history homework short text is a text for expressing the homework question answer by the target student. The historical job information is compared with the job response information, so that whether the job response information is the analyzable job response information is determined according to the comparison result of the historical job information and the job response information. If yes, the homework response information of the target students is input into a preset comprehensive evaluation model so as to sequentially match a plurality of analysis evaluation texts of the target students in a preset evaluation dimension set. Wherein evaluating the set of dimensions comprises: student work accuracy dimension, work stability dimension, work time efficiency management dimension, class work accuracy dimension. And generating an operation analysis report of corresponding statistical period according to each analysis and evaluation text and the evaluation word set corresponding to the target student. The set of evaluation mood words includes mood words for matching the expression job analysis report.

In one implementation of the application, the reference index of each dimension of the target student in the statistical period is determined through the comprehensive evaluation model. The dimension reference index includes: an operation overall analysis index, an operation amount analysis index, an operation accuracy analysis index, an analysis index of each teaching aid/test paper, an error question analysis index and a question analysis index of the superior and inferior conditions. Based on the reference index of each dimension, each analysis and evaluation text is generated. Generating an operation analysis report of a corresponding statistical period according to each analysis and evaluation text and an evaluation word set corresponding to a target student, wherein the operation analysis report comprises the following specific steps:

and matching corresponding language information sets according to the personal information of the target students. The personal information includes: age, sex, hobbies. The mood information set includes a plurality of emotional mood words.

And matching the emotion mood words of the mood information set with the language colors of the analysis evaluation texts to remove the emotion mood words with the unmatched language colors from the mood information set so as to obtain an evaluation mood word set.

The preset decoder is trained by evaluating the vocabulary of intonation. The decoder is configured to output text corresponding to the set of evaluated intonation words.

And inputting each analysis and evaluation text into a preset encoder to obtain corresponding encoding vectors. And inputting the encoded vector into a trained decoder to generate a job analysis report.

In one implementation manner of the method, fluctuation variation of the work average correct rate of the target students in the statistical period and the work average correct rate of the last statistical period of the ring ratio is determined through a first preset algorithm. Wherein the first preset algorithm comprises:

wherein,for the average accuracy of the job, N is the number of submitted jobs, N is a natural number, ++>For the target student at H _i Job accuracy of the secondary commit job.

Wherein DeltaSo isFor the fluctuation variable, +.>Job average correctness class ranking, which may correspond to the current statistical period, +.>Job average correctness class ranks may correspond to a period immediately preceding the current statistical period. And taking the operation average accuracy and fluctuation variation as operation integral analysis indexes.

In one implementation manner of the method, the number of submitted homework times, the number of homework exercises and the actual homework amount mean value of the target students are compared with the corresponding class homework amount mean value through a second preset algorithm, so that the corresponding first comparison result is used as an homework amount analysis index. Wherein the second preset algorithm comprises:

N-hteNum＝{H ₁ ，H ₂ ，......，H _N }

wherein N-hteNum is the number of times of submitting the operation;

N-hteSubExerciseNum＝{E ₁ ，E ₂ ，......，E _M }

wherein N-hteStubExerciseNum is the number of the operation problems;

Wherein N-wrungHteSubExerciseNum is the number of wrong questions, M is the number of questions of wrong questions, and is less than or equal to M;

wherein, T-teacherSuggestTime is a set of recommended duration of a single operation of a teacher,to be at H ₁ The recommended duration of the secondary job;

wherein, T-stuAugAnsweTime is the actual answer time set of each job,for the actual answer of the 1 st job>For the actual answer of the 2 nd job +.>The actual answer time of the nth operation is the actual answer time;

wherein, stuAvgAnsweTime is the average duration of actual response in the student homework statistics period, N is the number of times the student submitted homework, which is less than or equal to N;

wherein,longitudinal fluctuation value for working time effect->For the actual answer average duration of the current statistical period,/-, for the current statistical period>The actual answer average time length of the last statistical period is the same as the actual answer average time length of the last statistical period;

wherein,time for the work aging transverse contrast value ^student For the actual answer average duration and Time of the current statistical period ^class And the actual answering average time length of the class students in the current statistical period is equal.

In one implementation manner of the method, the work score stability analysis value of the target student is calculated based on the work average correct rate and the class average correct rate of the target student through a third preset algorithm. And taking a matching result of the work average accuracy and the class average accuracy and a work result stability analysis value as work accuracy analysis indexes. Wherein the third preset algorithm comprises:

Wherein,the difference value between the average accuracy rate of the homework for the target students and the average accuracy rate of the class students is calculated; s is S ^class Average correct rate for class students;

wherein σS ^student The method comprises the steps of counting a stability analysis value of the homework score in a period for students;

wherein G (S_R) is a job fluctuation level characterization value, beta ₀ 、β ₁ 、β ₂ Is a preset parameter.

In one implementation of the present application, the accuracy of the teaching aid/test paper in the job response information is determined by a fourth preset algorithm. And comparing the correct rate of the teaching aid/test paper, the average correct rate of the operation and the average correct rate of the class, and taking the corresponding second comparison result as analysis indexes of the teaching aid/test paper. And respectively comparing the question accuracy with the question class accuracy and the grade question accuracy according to the question serial numbers of the teaching aid/test paper in the operation response information by a fourth preset algorithm, and taking the corresponding third comparison result as a wrong question analysis index. Determining the target students through a fourth preset algorithmThe quality and inferiority questions are analysis indexes of the quality and inferiority questions. The merits and merits include: master better questions, error-prone questions and focus on questions. The fourth preset algorithm comprises the following steps: es = isRight; wherein Es represents whether the result is correct, including correct true and False, isRight represents whether the result is correct; grad, egs, represents the average accuracy of individual questions in the teaching aid/test paper.

In one implementation of the present application, the mastering of better questions satisfies that the questions are answered correctly and the corresponding rank questions accuracy is less than a first preset threshold. The error-prone questions meet that the corresponding grade question accuracy rate is less than a second preset threshold. The important attention questions meet the question answering errors, and the corresponding grade question accuracy rate is larger than a third preset threshold.

In one implementation manner of the method, a student work accuracy rate sub-model and a work stability sub-model are determined according to a first preset algorithm and a third preset algorithm so as to match analysis evaluation texts corresponding to student work accuracy rate dimensions and work stability dimensions. And determining an operation time efficiency management sub-model and an operation accuracy sub-model according to the first preset algorithm and the second preset algorithm so as to match the analysis evaluation text corresponding to the operation time efficiency management dimension and the class operation accuracy dimension. And matching the corresponding analysis evaluation text of the target students according to a fourth preset algorithm and the analysis indexes of the teaching aid/test paper and the analysis indexes of the questions of the advantages and disadvantages. And matching preset dialects corresponding to the dimensions according to the reference indexes of the dimensions. The speech is the analysis evaluation text of the preset emotion color.

In one implementation of the present application, the assignment analysis report is sent to the user terminal, so that the assignment analysis report corresponding to the target student is displayed at the user terminal. The usage object of the user terminal at least comprises one or more of the following roles: parents, students, teachers.

In one implementation of the application, determining a user terminal accessing a link corresponding to the assignment analysis report of the target student and an access frequency thereof; according to the access frequency, determining the attention weight of the corresponding user terminal; determining the preference weight of the corresponding user terminal according to the access content of the job analysis report corresponding to the access; the preference weights comprise student work correct rate dimension preference weights, work stability dimension preference weights, work timeliness management dimension preference weights and class work correct rate dimension preference weights; and determining the user terminal for sending the job analysis report and the display content of the job analysis report from a plurality of user terminals based on the attention weight and the preference weight.

On the other hand, the embodiment of the application also provides a job analysis device, which comprises:

The acquisition module acquires the historical homework information of the target students and the current homework response information of the target students. The history job information comprises the history job answering time and the history job short text of each question type. The history homework short text is a text for expressing the homework question answer by the target student.

And the comparison module is used for comparing the historical job information with the job response information so as to determine whether the job response information is analyzable job response information according to the comparison result of the historical job information and the job response information.

The input module is used for inputting the homework answering information into a preset comprehensive evaluation model so as to sequentially match a plurality of analysis evaluation texts of the target students in a preset evaluation dimension set. Wherein evaluating the set of dimensions comprises: student work accuracy dimension, work stability dimension, work time efficiency management dimension, class work accuracy dimension.

And the generation module is used for generating an operation analysis report with corresponding statistical period according to each analysis and evaluation text and the evaluation word set corresponding to the target student. The set of evaluation mood words includes mood words for matching the expression job analysis report.

According to the technical scheme, the homework answering information of the target students is determined to be matched with the historical homework answering information, then the homework answering information of the students is processed by using the comprehensive evaluation model, the multi-contrast dimension and angle speaking text is obtained, and an homework analysis report is generated according to the evaluation word set corresponding to the target students. And furthermore, objective evaluation and personalized display of the homework analysis report can be carried out on the procedural homework of the students, so that the students can resonate the homework analysis report, and the result guidance of the homework analysis report can be effectively understood. In addition, the report is generated in multiple contrast dimensions and multiple angles, so that knowledge holes of students can be effectively reflected through the homework report, and the learning performance of the students is effectively improved.

And 1) the overall work quality of the period is statistically analyzed and counted, the fluctuation condition of the work quality of the period is compared, and the correlation summary and suggestion of the overall work level of the period of the student are given; 2) Deeply analyzing the homework amount in the statistical period, assisting parents or students to master homework frequency and answering amount in the period, enabling the students to actually answer time, further analyzing homework timeliness of the students by combining the homework quality, summarizing homework characteristics, and matching the matched learning lifting suggestions according to the homework characteristics; 3) Analyzing the stability of the homework quality of students in different periods, transversely comparing the homework effect and the stability of the whole class, and longitudinally comparing whether the students are advanced or not; 4) Counting wrong questions, displaying wrong questions in a personal period, comparing class accuracy, locating and grasping weaker questions by observing the questions in different grasping regions, checking leakage and supplementing defects, and accurately reviewing.

Drawings

The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiments of the application and together with the description serve to explain the application and do not constitute an undue limitation to the application. In the drawings:

FIG. 1 is a flow chart of a job analysis method according to an embodiment of the present application;

Fig. 2 is a schematic structural diagram of a job analysis apparatus according to an embodiment of the present application.

Detailed Description

For the purposes, technical solutions and advantages of the present application, the technical solutions of the present application will be clearly and completely described below with reference to specific embodiments of the present application and corresponding drawings. It will be apparent that the described embodiments are only some, but not all, of the embodiments of the present application. 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.

The current statistics index of the student homework accuracy is not comprehensive enough, when the student homework is analyzed, the comparison dimension and the comparison angle are single, and when the student is evaluated, the result is heavier than the result and the process is lighter, so that the evaluation of the student is not objective. The content of the homework report is dead, and students cannot resonate the homework report, so that knowledge holes of the students cannot be effectively reflected through the homework report, and the learning performance of the students is improved.

Based on the above, the embodiment of the application provides a homework analysis method and device, which are used for solving the technical problems that the evaluation of the current student procedural homework report is not objective, students cannot resonate the homework report, knowledge holes of the students cannot be effectively reflected through the homework report, and the learning performance of the students is improved.

Various embodiments of the present application are described in detail below with reference to the accompanying drawings.

The embodiment of the application provides a job analysis method, as shown in fig. 1, the method may include steps S101 to S104:

s101, the server acquires historical homework information of the target students and current homework response information of the target students.

The history job information comprises the history job answering time and the history job short text of each question type. The history homework short text is a text for expressing the homework question answer by the target student.

The server is an execution subject of the job analysis method, and is merely exemplary, and the execution subject is not limited to the server, and the present application is not particularly limited thereto.

The homework response information comprises response homework submitted to the server by the target students, and comprises the following steps: task questions and task response determination results. The job answer determination result may include: correct or incorrect determination results, job title scores, etc.

The historical homework information of the target students is the historical homework answering time length of the homework of each question type in the historical homework answering of the target students, and the historical homework short text of answer description is used when the homework is answered. Through the historical homework information, the homework answering habit of the target students can be reflected, for example, one choice question is selected, the duration is generally 5-20 seconds, one subjective question is selected, and the homework answering habit of the target students can be reflected by the language order, the common words and the like of the answer description of the target students. And taking the homework answering habits of the target students as historical homework information.

The current job response information may be job response information within a certain statistical period, for example, one week or one month.

S102, the server compares the historical job information with the job response information to determine whether the job response information is analyzable job response information according to the comparison result of the historical job information and the job response information.

In the embodiment of the application, the server can compare the historical operation information with the operation response information, wherein the comparison can be to compare the operation response time length in the current operation response information with the historical operation response time length consistent with the question type; the subjective question answering text of the question type A in the current job answering information can be compared with the history job short text. The comparison may be to calculate the cosine similarity of the comparison content, and determine that the job response information is the analyzable job response information when the cosine similarity is greater than a predetermined value, and if the cosine similarity is less than the predetermined value, generate alarm information and send the alarm information to a corresponding management terminal, where the management terminal may be a mobile phone of a teacher, a computer of a educational administration part, and other devices. The warning information is used for reminding the relevant personnel of the management terminal, the homework response information of the target student is inconsistent with the historical homework information, and the management terminal personnel selects whether to continue homework analysis or update the selection range of the historical homework information.

In one example, the historical homework information of the target student may be bad, the homework answer and the achievement of the target student may not match with the historical homework information gradually, and then the server generates alarm information and the management terminal adjusts the historical homework information.

In addition, the subjective question answering text of the question type A in the current operation answering information is compared with the history operation short text to obtain a comparison result, and the subjective question answering text and the history operation short text can be subjected to word segmentation processing respectively to obtain two groups of text groups to be compared. And then according to the word sequence of the two answer texts, comparing the similarity which is met by the words and the punctuations in the text group to be compared, wherein the similarity comprises but is not limited to the similarity of the frequencies of the words and the punctuations at the positions of the texts in which the words and the punctuations are positioned.

S103, the server inputs the homework response information into a preset comprehensive evaluation model under the condition that the homework response information is determined to be the analyzable homework response information, so that a plurality of analysis evaluation texts of the target students are sequentially matched in a preset evaluation dimension set.

Wherein evaluating the set of dimensions comprises: student job accuracy dimension, job stability dimension, job time-efficiency management dimension, class job accuracy dimension (class job average accuracy dimension).

In the embodiment of the application, the homework response information of the target students is input into a preset comprehensive evaluation model so as to sequentially match a plurality of analysis evaluation texts of the target students in a preset evaluation dimension set, and the method specifically comprises the following steps:

the server can determine the reference index of each dimension of the target students in the statistical period through the comprehensive evaluation model. The dimension reference index includes: an operation overall analysis index, an operation amount analysis index, an operation accuracy analysis index, an analysis index of each teaching aid/test paper, an error question analysis index and a question analysis index of the superior and inferior conditions. Then, each analysis evaluation text is matched based on each dimension reference index.

The method for analyzing the homework overall in the student period to obtain homework overall analysis indexes specifically comprises the following steps:

the student can complete the guide language analysis of the homework submitted by the student in a statistical period, for example, one week is a statistical period, and the homework answering information of the homework answering completed by the student in the period is collected.

And then, determining the work average correct rate of the target students in the current statistical period and the fluctuation variation of the work average correct rate of the last statistical period of the ring ratio through a first preset algorithm.

The first preset algorithm can count the periodic student homework score conditions: and counting the average score of all homeworks submitted by students in the period. The formula is as follows:

Wherein,for the average accuracy of the job, N is the number of times of submitting the job, N is a natural number, ++>For the target student at H _i Job accuracy of the secondary commit job.

And can calculate the fluctuation condition of the job achievement: the fluctuation variation of the average accuracy rate of the work of the student ring compared with the last statistical period is as follows:

wherein DeltaSo is the fluctuation variation,job average correctness class ranking, which may correspond to the current statistical period, +.>Job average correctness class ranks may correspond to a period immediately preceding the current statistical period.

Then, the server uses the job average accuracy and the fluctuation variance as the overall analysis index of the job.

According to the method and the device, the preset conversation corresponding to each dimension can be matched according to the reference index of each dimension. The speech is the analysis evaluation text of the preset emotion color.

That is, after obtaining the overall analysis index of the job, the server can use the index to match the speech related to the overall analysis of the job, and further obtain the analysis evaluation text that constitutes a part of the job analysis report. The specific matching result is as follows:

first outcome-advancement: the horizontal ring of the student in the period school is improved compared with the upper period;

1. and delta So is less than 0, the average accuracy of the periodic work is greater than or equal to the average accuracy of the periodic work, and the student performance progress is judged to be matched with the relatively positive speaking skill.

2. So/Ns < = 1/3, where So is the class ranking of the current statistical period of students, ns is the number of people in class, and for the situation that students do not submit homework in the first statistical period or the last statistical period, the average accuracy of homework submitted in the current period is ranked in class. If the rank is in the top 1/3, the job is considered to be higher in completion, matching the corresponding conversation.

Second result-leveling: the student's periodic academic horizontal ring is relatively leveled with the upper period;

and delta So is more than or equal to 0 and less than or equal to So0, the ranking is not floated downwards and exceeds a ranking floating threshold So0, such as 3, and the ranking is regarded as score leveling and is matched with a corresponding conversation.

The So/Ns is more than 1/3 and less than or equal to 2/3, and aiming at the situation that students do not submit homework in the first statistical period or the last statistical period, the personal rank of the students is 1/3 before the class, the homework completion degree is higher, and the corresponding conversation is matched.

Third outcome-step back: the student's periodic academic horizontal ring is lowered compared with the upper period;

Δso > So0, the student rank falls beyond the rank floating threshold So0 by 3 if the cycle of the current cycle academic level cycle is compared with the cycle, and the cycle of the current cycle is determined to learn to return steps and match the corresponding conversation.

So/Ns >2/3, for the first statistical period or the last statistical period when students did not submit homework, the personal rank of students is 1/3 after the class, the achievement is considered to be poor, and the corresponding conversation is matched.

Through the first preset algorithm, the corresponding voice operation under the whole operation analysis index can be matched, and then analysis evaluation text is generated.

The work amount analysis index is obtained by analyzing the work amount of students, and specifically comprises the following steps:

and comparing the number of submitted homework times, the number of homework exercises and the actual homework amount mean value of the target students with the corresponding class homework amount mean value through a second preset algorithm, so that the corresponding first comparison result is used as an homework amount analysis index.

The calculation process of the second preset algorithm is specifically as follows:

the number of operations: the students finish the submitting homework in the selected period;

N-hteNum＝{H ₁ ，H ₂ ，......，H _N }

wherein N-hteNum is the number of operations.

Number of exercises: the operation in the period comprises M problems;

N-hteSubExerciseNum＝{E ₁ ，E ₂ ，......，E _M }

wherein N-hteStubExerciseNum is the number of task topics.

Number of wrong questions: collecting the number of wrong questions in the homework after the students complete the answering;

wherein N-wrungHteSubExerciseNum is the number of wrong questions, M is the number of questions of wrong questions, and is less than or equal to M.

Calculation of single operation time:

1. the teacher suggests the duration of a single operation: the teacher arranges the recommended duration of each operation in the period;

wherein, T-teacherSuggestTime is a set of recommended duration of a single operation of a teacher, To be at H ₁ The recommended duration of the secondary job.

2. The actual answering time of the single homework of the student: the actual answering time for the students to complete the homework in the period;

wherein, T-stuAugAnsweTime is the actual answer time length set of each job,the actual answering time length (actual answering time) of the 1 st operation; />For the actual answer of the 2 nd job +.>The actual answer time of the nth job is used.

3. The average length of actual answers in the student period: an average value of actual answering time lengths of all jobs submitted by students in a period;

wherein, stuAvgAnsweTime is the average length of actual response in the student work statistics period, N is the number of times the student submitted the work, which is less than or equal to N.

4. Longitudinal fluctuation condition of operation aging: the student ring fluctuates in comparison with the actual answering average duration of the last statistical period;

wherein,longitudinal fluctuation value for working time effect->For the actual answer average duration of the current statistical period,/-, for the current statistical period>The length of the actual answer average for the last statistical period.

5. Transverse comparison condition of operation aging: the students compare the average time difference of actual answers of the students in the class in the same period;

wherein,time for the work aging transverse contrast value ^student For the actual answer average duration and Time of the current statistical period ^class And (5) the average time length average value of actual answers of class students in the current statistical period.

The method and the device compare the number of submitted homework times, the number of homework exercises and the actual answering time of the target students with the corresponding class homework volume average value, namely compare the number of submitted homework times and the number of homework exercises of the target students with the average actual answering time of the class students, and obtain the comparison result of the homework accuracy rate of the students and the class homework accuracy rate according to the number of submitted homework times and the number of homework exercises of the target students, and serve as a first comparison result, namely an homework volume analysis index.

The specific matching result of the workload analysis index is as follows:

the time is short: the actual answering time of the students is smaller than the average actual answering time of the students in the class;

1. the average job accuracy of the submitted homework of the students is in the first third of the average job accuracy of the class, the time for the statistical periodic homework of the target students is determined to be short, the homework accuracy is high, and the corresponding conversation is matched;

2. the average job accuracy of the submitted homework of the students is between one third and two thirds of the average job accuracy of the class, the statistical period homework response time of the target students is determined to be short, the homework accuracy is general, and the homework accuracy is matched with the corresponding speaking;

3. The average job accuracy of the submitted homework of the students is in the last third of the average job accuracy of the class, the time for the statistical period homework of the target students is determined to be short, the homework accuracy is low, the homework attitude is more traumatology, the homework is suggested to be carefully completed, and the corresponding speaking operation is matched.

Time is medium: the difference value between the actual answering time of the student and the average actual answering time of the class students is within the preset time, for example, 5 minutes, and the student homework time is considered to be medium;

1. the average accuracy of the homework of students is in the first third of the class, the statistical period homework time of the target students is determined to be medium, the homework accuracy is high, and the corresponding conversation is matched;

2. the average work accuracy of students is positioned at the middle position of a class (namely, so/Ns is more than 1/3 and less than or equal to 2/3), the statistical period work time of the target students is determined to be medium, the work accuracy is general, and the corresponding conversation is matched;

3. the student accuracy is positioned at the rear third of the class, the target students are determined to count the period homework time, the homework accuracy is low, and the corresponding conversation is matched.

The time is long: when the student actually answers, the difference value between the student individual actually answers and the average time length of the class student actually answers exceeds a preset time, for example, 5 minutes, and the student homework time is considered to be long;

1. In this case, the server may match the session to determine that the student needs to pay attention to the duration of the job to finish the job, and adjust the session method.

There is no time: counting the duration of the homework submitted by the students without collecting the homework;

1. the student accuracy is in the first third of the class, and the job accuracy is high, and the corresponding conversation under the condition of no time is matched;

2. the student accuracy is positioned at the middle position of the class (namely, so/Ns is more than 1/3 and less than or equal to 2/3), the average accuracy of the target student homework is determined to be general, and the corresponding conversation is matched under the condition of no time;

3. the student accuracy is positioned at the rear third of the class, the target student homework accuracy is determined to be lower, and the corresponding conversation is matched under the condition of no time.

Through the second preset algorithm, the corresponding voice operation under the workload analysis index can be matched, and then analysis evaluation text is generated.

The method for analyzing the work accuracy rate in the student period comprises the following steps of:

and calculating the work score stability analysis value of the target student based on the work average correct rate and the class average correct rate of the target student through a third preset algorithm. And the matching result of the work average correct rate and the class average correct rate and the work score stability analysis value are used as work correct rate analysis indexes.

Specifically, the third preset algorithm includes:

1. counting the student homework score conditions in the period:

2. longitudinal fluctuation condition of work score: the average accuracy rate of the homework of the student personal ring is fluctuated compared with the last statistical period;

3. transverse comparison condition of operation score: the students compare the average correct rate difference value of class students in the same statistical period;

wherein,the difference value between the average accuracy rate of the homework for the target students and the average accuracy rate of the class students is calculated; s is S ^class Average correct rate for class students.

4. Performing score stability analysis;

5. and (3) comprehensive model: determining a student work fluctuation level characterization model by the student work actual answering effect:

And the third preset algorithm can judge the result, and takes the matching result of the operation average correct rate and the class average correct rate and the operation score stability analysis value as the operation correct rate analysis index to match the corresponding operation, which is specifically as follows.

1. The So/Ns is less than or equal to 1/3, and the & Ls is less than or equal to Lc, wherein the & is a logic condition AND, ls is a ranking dispersion obtained by ranking the student personal at each submitted homework score, and Lc is a class ranking dispersion; the students submit homework in the last statistical period, the upper reaches of the student results, the personal ranking is positioned at the position 1/3 of the position before the class, the personal dispersion of the students is less than or equal to the class dispersion average value, the student results in the last statistical period are considered to be stable at the upper reaches, and the student results are matched with the corresponding conversation.

2. And So/Ns is less than or equal to 1/3 &ls > Lc, students submit homework in the cycle than the last statistical period, the personal rank is positioned at the position 1/3 of the front class, the personal dispersion of the students is greater than the average value of the class dispersion, and the student periodic performance is considered to be stable in the middle and upstream, but the fluctuation is large and matches the corresponding call.

3. So/Ns is less than or equal to 1/3& & Ls is less than or equal to Lc ≡Lm0, wherein Lm is the difference between the personal dispersion of the student and the personal dispersion of the last statistical period, and Lm0 is a dispersion difference threshold, such as 2; the personal ranking is positioned at the position 1/3 of the front of the class, the personal dispersion of the students is less than or equal to the average value of the dispersion of the class, the dispersion is equal to or lower than the average value of the dispersion of the classes, the periodic performance of the students is considered to be positioned at the upper reaches, and the corresponding conversation is matched.

4. The So/Ns is less than or equal to 1/3 &lsis less than or equal to Lc &lm > Lm0, the student score is upstream, the personal rank is positioned at the position 1/3 of the front class, the personal dispersion of the student is less than or equal to the average value of the class dispersion, the dispersion is improved compared with the upper week, the score in the period of the student is considered to be upstream, but the score fluctuation amplitude is larger than the upper period, and the score is matched with the corresponding conversation.

5. The So/Ns is less than or equal to 1/3 &ls > Lc & Lm < 0, the student score is upstream, the personal rank is positioned at the position 1/3 of the front class, the student personal dispersion is greater than the class dispersion mean value, the dispersion is reduced compared with the upper week, the student cycle score is considered to be upstream, but the score fluctuation is fluctuated, and the student score is matched with the corresponding conversation.

6. The So/Ns is less than or equal to 1/3 &ls > Lc &Lmis more than or equal to 0, the upper stream of student results is provided with a personal rank at the position 1/3 of the front of the class, the personal dispersion of the students is greater than the average value of the dispersion of the class, the dispersion is equal to or higher than the upper week, the result of the cycle of the student is regarded as being at the upper stream, but the result fluctuates, and the result is matched with the corresponding conversation.

7. The So/Ns is more than 1/3 and less than or equal to 2/3 and the Ls is more than or equal to Lc, the student results are moderate, the personal rank is positioned at the middle position (1/3 to 2/3) of the class, the personal dispersion of the student is less than or equal to the average value of the class dispersion, the result of the cycle of the student is considered to be positioned at the middle position, the result is relatively stable, and the student is matched with the corresponding conversation.

8. The So/Ns is more than 1/3 and less than or equal to 2/3 and the Ls is more than Lc, the student score is medium, the personal rank is positioned at the middle position of the class (1/3 to 2/3), the personal dispersion of the student is more than the average value of the class dispersion, the score of the cycle of the student is considered to be positioned at the middle position, the score fluctuates and is matched with the corresponding conversation.

9. And So/Ns >2/3 &lsis less than or equal to Lc, the student score is lower, the personal rank is positioned at the position of 1/3 after the class, the personal dispersion of the student is less than or equal to the class dispersion mean value, the student periodic score is considered to be positioned at the position of the back, the score is relatively stable, and the student periodic score is matched with the corresponding conversation.

10. And So/Ns >2/3 &ls > Lc, the student score is lower, the personal rank is positioned at the post-class 1/3 position, the personal dispersion of the student is > class dispersion mean value, the student periodic score is considered to be positioned at the post-class position, the score fluctuates and is matched with the corresponding conversation.

In addition, the third preset algorithm can also determine the homework with lower accuracy and the homework name needing specific attention according to the average accuracy comparison analysis of individuals and classes of homework in each student counting period, and remind students of focusing on the homework.

The method and the device for analyzing the teaching aid/test paper in the student counting period obtain teaching aid/test paper analysis indexes, wrong question analysis indexes and good and bad question analysis indexes by analyzing the teaching aid/test paper, wrong question analysis indexes and good and bad question analysis indexes in the student counting period, and specifically comprise the following steps:

and determining the accuracy of the teaching aid/test paper in the operation response information through a fourth preset algorithm. And comparing the correct rate of the teaching aid/test paper, the average correct rate of the operation and the average correct rate of the class, and taking the corresponding second comparison result as analysis indexes of the teaching aid/test paper.

And respectively comparing the question accuracy with the question class accuracy and the grade question accuracy according to the question serial numbers of the teaching aid/test paper in the operation response information by a fourth preset algorithm, and taking the corresponding third comparison result as a wrong question analysis index.

And determining the questions of the good and bad vigor of the target students as the analysis index of the questions of the good and bad vigor through a fourth preset algorithm. The merits and merits include: master better questions, error-prone questions and focus on questions. The method comprises the steps of obtaining a first preset threshold value, wherein the first preset threshold value is smaller than the first preset threshold value, and the first preset threshold value is smaller than the second preset threshold value. The error-prone questions meet that the corresponding grade question accuracy rate is less than a second preset threshold. The important attention questions meet the question answering errors, and the corresponding grade question accuracy rate is larger than a third preset threshold. The first preset threshold, the second preset threshold and the third preset threshold are preset, which is not particularly limited in this application. For example, the first preset threshold is 30%, the second preset threshold is 50%, and the third preset threshold is 70%.

The analysis of each teaching aid/test paper specifically comprises the following steps:

and a fourth preset algorithm is adopted, the accuracy of teaching assistance/test paper in the teacher carefully chosen work is determined according to the source of the teacher carefully chosen work, the personal overall accuracy of the questions of the students in the teaching assistance/test paper is compared with the class average accuracy, and a second comparison result is obtained, so that the mastering condition of the students on different teaching assistance/test paper is determined, the mastering degree of the different teaching assistance/test paper is combined, the corresponding speaking is matched, and the speaking is the speaking recommended for the student work.

The error question analysis specifically comprises the following steps:

and determining the corresponding question serial numbers of the teaching aids/test papers submitted every day in the statistical period, the correct conditions corresponding to the questions, the correct rate and the class correct rate of the questions through a fourth preset algorithm, comparing the correct rates with the correct conditions corresponding to the questions at class level, comparing the correct rates with the difference value of the class correct rates of the questions, and matching the corresponding speech skills according to the compared third comparison result.

The analysis of the superior and inferior questions comprises:

and determining better mastering questions, error-prone questions and important attention questions in a statistical period through a fourth preset algorithm. Respectively matching the corresponding speech technologies; the problems are classified into three categories of better mastering problems, error-prone problems and important attention problems, and students can fully know own dominant problems and insufficient problems, so that the training and consolidation are enhanced.

The error question statistical analysis is specifically as follows:

1. whether a student holds a single question: es = isRight; es indicates whether the result is correct, including correct true and False, isRight indicates whether it is correct.

2. Average correct rate of individual topics for students:grad, egs, represents the average accuracy of individual questions in the teaching aid/test paper.

According to the calculation result of the fourth preset algorithm, the result judgment is carried out, and the corresponding call operation is matched, specifically as follows:

1. master the better questions: es=true and Egs is less than a first preset threshold, and the student individual question answer pairs and the grade average correct rate are less than the first preset threshold;

2. error-prone problem: egs is less than a second preset threshold, i.e., the average accuracy of the grade is less than, for example, 50%;

3. the key points of interest are: es=false and Egs is greater than a third preset threshold, i.e. the result of a single question by the student is wrong, the question rank average correct rate is greater than a third preset threshold, e.g. 70%.

In addition, the application server matches each analysis and evaluation text based on each dimension reference index, and further comprises:

and the server determines a student work accuracy rate sub-model and a work stability sub-model according to the first preset algorithm and the third preset algorithm so as to match analysis and evaluation texts corresponding to the student work accuracy rate dimension and the work stability dimension.

For example, the method is obtained according to a student work accuracy rate sub-model and a work stability sub-model:

1. g (S_R) is more than or equal to 0, generating analysis evaluation text which is good in comprehensive performance, stable in performance and continuously kept for the statistics period of the target students;

2. g (S_R) is less than 0, the comprehensive level of the score of the statistics period of the target students needs to be improved, the score fluctuation is large, and the analysis evaluation text of the basic exercise is suggested to be strengthened.

And determining an operation time efficiency management sub-model and an operation accuracy sub-model according to the first preset algorithm and the second preset algorithm so as to match the analysis evaluation text corresponding to the operation time efficiency management dimension and the class operation accuracy dimension. And matching the corresponding analysis evaluation text of the target students according to a fourth preset algorithm and the analysis indexes of the teaching aid/test paper and the analysis indexes of the questions of the advantages and disadvantages.

Specifically, 1, G (Time) is equal to or more than 0,matching analysis evaluation text for prompting students to pay attention to working hours in answering process, wherein

Wherein beta is ₃ 、β ₄ 、β ₅ Is a preset parameter.

2、G(Time)≥0，Matching analysis evaluation texts which prompt students that the work completion degree is not high and the work quality needs to be noted;

3、G(Time)＜0，the matching exaggeration student has high work completion degree and less time consumption;

4、G(Time)＜0，And matching and prompting students to work attitude right, and carefully completing analysis and evaluation texts of the work.

S104, the server generates an operation analysis report with corresponding statistical period according to each analysis and evaluation text and the evaluation word set corresponding to the target student.

The evaluation of the population of mood words is used to match the mood words of the expression job analysis report.

According to each analysis and evaluation text and an evaluation word set corresponding to a target student, generating an operation analysis report with a corresponding statistical period, wherein the operation analysis report specifically comprises the following steps:

the server matches corresponding sets of mood information according to personal information of the target students. The personal information includes: age, sex, hobbies. The personal information may be information entered by a student or teacher in a preset database. For example, the age is 10 years, sex men, hobbies are basketball, and then the matched set of mood information can correspond to the mood of a ball star; age 13, gender female, hobbies singing, then the matching mood information may correspond to a singer. The term information set may be obtained by training a predetermined neural network model to obtain a plurality of dialogue samples (the dialogue samples include terms including speech terms, term words, and term words including language colors). The mood information set includes a plurality of emotional mood words. Then, the server matches each emotion word of the mood information set with each emotion word of the analysis evaluation text to remove emotion words with unmatched emotion colors in the mood information set so as to obtain an evaluation mood word set.

Language colors can be understood as words of different emotional degrees, e.g. good, very good, better, words belonging to different language colors.

The server then trains the pre-set decoder by evaluating the vocabulary of the mood words. The decoder is configured to output text corresponding to the set of evaluated intonation words.

The decoder is obtained through training a plurality of emotion words in the evaluation word set, and during training, the server can input a plurality of text samples and emotion words in the evaluation word set corresponding to each text sample, so that the decoder is trained, and the text samples and the emotion words can be in one-to-one or many-to-one relation.

And finally, inputting each analysis and evaluation text into a preset encoder to obtain corresponding encoding vectors. And inputting the encoded vector into a trained decoder to generate a job analysis report.

The decoder sequentially adds the decoded analysis evaluation texts to the job analysis report according to the order of the input analysis evaluation texts.

Through the scheme, the technical problems that evaluation of the current student procedural homework report is not objective, students cannot resonate the homework report, knowledge holes of the students cannot be effectively reflected through the homework report, and the learning performance of the students is improved can be solved. Objective evaluation can be carried out on the procedural homework of students, so that the students can resonate the homework analysis report, and the result guidance of the homework analysis report can be effectively understood.

Furthermore, it is also possible to realize: 1) The overall work quality of the period is counted, the fluctuation condition of the work quality of the period is compared, and the correlation summary and suggestion of the overall work level of the period of the student are given; 2) Deeply analyzing the homework amount in the statistical period, assisting parents or students to master homework frequency and answering amount in the period, enabling the students to actually answer time, further analyzing homework timeliness of the students by combining the homework quality, summarizing homework characteristics, and matching the matched learning lifting suggestions according to the homework characteristics; 3) Analyzing the stability of the homework quality of students in different periods, and transversely comparing the homework effect and the stability of the whole class, and longitudinally and per se advancing or not; 4) Counting wrong questions, displaying wrong questions in a personal period, comparing class accuracy, locating and grasping weaker questions by observing the questions in different grasping regions, checking leakage and supplementing defects, and accurately reviewing.

And furthermore, the procedural homework of the students can be objectively evaluated, and reports can be generated in multiple contrast dimensions and multiple angles, so that knowledge holes of the students can be reflected through the homework reports, and the learning performance of the students can be effectively improved.

In addition, after the server generates the job analysis report of the corresponding statistical period, the method further comprises:

And sending the homework analysis report to the user terminal so as to display the homework analysis report corresponding to the target student at the user terminal. The usage object of the user terminal at least comprises one or more of the following roles: parents, students, teachers.

The user terminal can be a device such as a mobile phone, a computer and the like of parents, students or teachers, and the type of the device is not particularly limited.

In the embodiment of the application, the server can also display the homework analysis report in a personalized manner, and as the roles related to the target students are numerous and the attention dimensions of the homework analysis report are different, the application can display the homework analysis report in a personalized manner through the following embodiments.

Specifically, the server determines the user terminal accessing the link corresponding to the assignment analysis report of the target student and the access frequency thereof. And determining the attention weight of the corresponding user terminal according to the access frequency. And determining the preference weight of the corresponding user terminal according to the access content of the access corresponding job analysis report. The preference weights include student work accuracy dimension preference weights, work stability dimension preference weights, work timeliness management dimension preference weights, class work accuracy dimension preference weights. Based on the attention weight and the preference weight, a user terminal for transmitting the job analysis report and the display content of the job analysis report are determined from a plurality of user terminals.

That is, the attention weight of each user terminal to the target student is determined based on the frequency of connection of the user terminal to the presentation job analysis report, for example, the frequency is a, the attention weight is a, the access frequency is B, the attention weight is B, and the weight is greater than or equal to 0 and less than 1. And then accessing access records of contents in each dimension displayed by the homework analysis report of the target student according to the user terminal, for example, the homework accuracy dimension access times of the student are 2, and the content access times of the homework stability dimension are 1, wherein the homework accuracy dimension preference weight of the student is 0.5, and the homework stability dimension preference weight is 0.4. When the job analysis report is sent to the user terminal, the display sequence of the job analysis report in each dimension can be adjusted according to the ordering of the preference weights. In addition, the homework analysis report can be only sent to the user terminal with the attention weight of the target students being greater than the preset value, so that the homework analysis report can be effectively shared and personalized display of the homework analysis report is realized.

The embodiment of the application also provides a job analysis device, which comprises:

the acquisition module 201 acquires the historical homework information of the target student and the current homework response information thereof. The history job information comprises the history job answering time and the history job short text of each question type. The history homework short text is a text for expressing the homework question answer by the target student.

The comparison module 202 compares the history job information with the job response information to determine whether the job response information is analyzable job response information according to the comparison result of the history job information and the job response information.

The input module 203 is configured to input the homework response information of the target student into a preset comprehensive evaluation model, so as to sequentially match a plurality of analysis evaluation texts of the target student in a preset evaluation dimension set. Wherein evaluating the set of dimensions comprises: student work accuracy dimension, work stability dimension, work time efficiency management dimension, class work accuracy dimension.

And the generating module 204 is used for generating a job analysis report of a corresponding statistical period according to each analysis evaluation text.

The input module 203 is specifically configured to:

and determining the reference indexes of each dimension of the target students in the statistical period through the comprehensive evaluation model. The dimension reference index includes: an operation overall analysis index, an operation amount analysis index, an operation accuracy analysis index, an analysis index of each teaching aid/test paper, an error question analysis index and a question analysis index of the superior and inferior conditions.

Based on the reference index of each dimension, each analysis and evaluation text is generated.

The generating module 204 is specifically configured to:

The input module 203 is specifically configured to:

and determining the work average correct rate of the target students in the statistical period and the fluctuation variation of the work average correct rate of the last statistical period of the ring ratio through a first preset algorithm. Wherein the first preset algorithm comprises:

Wherein DeltaSo isFor the fluctuation variable, +. >Job average correctness class ranking, which may correspond to the current statistical period, +.>May correspond to the current statistics weekJob average correctness class rank for the last cycle of the period. And taking the operation average accuracy and fluctuation variation as operation integral analysis indexes.

The input module 203 is specifically configured to:

and comparing the number of submitted homework times, the number of homework exercises and the actual homework amount mean value of the target students with the corresponding class homework amount mean value through a second preset algorithm, so that the corresponding first comparison result is used as an homework amount analysis index. Wherein the second preset algorithm comprises:

N-hteNum＝{H ₁ ，H ₂ ，......，H _N }

wherein N-hteNum is the number of times of submitting the operation;

N-hteSubExerciseNum＝{E ₁ ，E ₂ ，......，E _M }

wherein N-hteStubExerciseNum is the number of the operation problems;

wherein, T-stuAugAnsweTime is the actual answer time set of each job,for the actual answer of the 1 st job>For the actual answer of the 2 nd job +.>The actual answer time of the nth operation is the actual answer time; / >

The input module 203 is specifically configured to:

and calculating the work score stability analysis value of the target student based on the work average correct rate and the class average correct rate of the target student through a third preset algorithm. And taking a matching result of the work average accuracy and the class average accuracy and a work result stability analysis value as work accuracy analysis indexes. Wherein the third preset algorithm comprises:

wherein σS ^student The stability analysis value of the operation score in the learning statistics period is obtained;

The input module 203 is specifically configured to:

And determining the questions of the good and bad vigor of the target students as the analysis index of the questions of the good and bad vigor through a fourth preset algorithm. The merits and merits include: master better questions, error-prone questions and focus on questions. The fourth preset algorithm comprises the following steps:

es = isRight; wherein Es represents whether the result is correct, including correct true and False, isRight represents whether the result is correct;

grad, egs, represents the average accuracy of individual questions in the teaching aid/test paper.

The input module 203 is specifically configured to:

the error-prone questions meet that the corresponding grade question accuracy rate is less than a second preset threshold.

The important attention questions meet the question answering errors, and the corresponding grade question accuracy rate is larger than a third preset threshold.

The input module 203 is specifically configured to:

and determining a student work accuracy rate sub-model and a work stability sub-model according to the first preset algorithm and the third preset algorithm so as to match analysis and evaluation texts corresponding to the student work accuracy rate dimension and the work stability dimension.

And matching preset dialects corresponding to the dimensions according to the reference indexes of the dimensions. The speech is the analysis evaluation text of the preset emotion color.

The generation module 204 is further capable of:

determining a user terminal accessing a link corresponding to the assignment analysis report of the target student and an access frequency thereof:

according to the access frequency, determining the attention weight of the corresponding user terminal;

determining the preference weight of the corresponding user terminal according to the access content of the job analysis report corresponding to the access; the preference weights comprise student work correct rate dimension preference weights, work stability dimension preference weights, work timeliness management dimension preference weights and class work correct rate dimension preference weights;

and determining the user terminal for sending the job analysis report and the display content of the job analysis report from a plurality of user terminals based on the attention weight and the preference weight.

All embodiments in the application are described in a progressive manner, and identical and similar parts of all embodiments are mutually referred, so that each embodiment mainly describes differences from other embodiments. In particular, for the device embodiments, since they are substantially similar to the method embodiments, the description is relatively simple, and reference is made to the description of the method embodiments in part.

The device and the method provided in the embodiments of the present application are in one-to-one correspondence, so that the device also has similar beneficial technical effects as the corresponding method, and since the beneficial technical effects of the method have been described in detail above, the beneficial technical effects of the device are not described here again.

It should also be noted that 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 foregoing is merely exemplary of the present application and is not intended to limit the present application. Various modifications and changes may be made to the present application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc. which are within the spirit and principles of the present application are intended to be included within the scope of the claims of the present application.

Claims

1. A method of job analysis, the method comprising:

acquiring historical homework information of a target student and current homework response information of the target student; the history job information comprises the history job answering time of each question type and the history job short text; the history homework short text is a text of the target student for expressing homework question answers;

comparing the historical job information with the job response information through presetting so as to determine whether the job response information is analyzable job response information according to a comparison result of the historical job information and the job response information;

if yes, inputting the homework response information into a preset comprehensive evaluation model so as to sequentially match a plurality of analysis evaluation texts of the target students in a preset evaluation dimension set; wherein the set of evaluation dimensions comprises: student homework accuracy dimension, homework stability dimension, homework time efficiency management dimension and class homework accuracy dimension;

generating an operation analysis report of a corresponding statistical period according to each analysis and evaluation text and an evaluation word set corresponding to the target student; the evaluation word set comprises emotion words used for matching and expressing the operation analysis report;

Inputting the homework response information into a preset comprehensive evaluation model so as to sequentially match a plurality of analysis evaluation texts of the target students in a preset evaluation dimension set, wherein the method specifically comprises the following steps of:

determining reference indexes of each dimension of the target students in the statistical period through the comprehensive evaluation model; the dimension reference index comprises: an operation overall analysis index, an operation amount analysis index, an operation accuracy analysis index, an analysis index of each teaching aid/test paper, an error question analysis index and a superior and inferior question analysis index;

generating each analysis and evaluation text based on each dimension reference index;

the method comprises the steps of determining, by the comprehensive evaluation model, reference indexes of each dimension of the target student in the statistical period, wherein the reference indexes comprise:

determining the work average correct rate of the target students in the statistical period and the fluctuation variation of the work average correct rate of the last statistical period of the ring ratio through a first preset algorithm;

wherein, the first preset algorithm comprises:

wherein,for the average accuracy of the job, N is the number of submitted jobs, N is a natural number, ++>For the target student at H _i The job accuracy of the secondary submitting job;

Wherein DeltaSo isFor the fluctuation variable, +.>Job average correctness class ranking, which may correspond to the current statistical period, +.>Job average correctness class ranking that may correspond to a period previous to the current statistical period;

taking the operation average correct rate and the fluctuation variation as the operation integral analysis index;

comparing the number of submitted homework times, the number of homework exercises and the actual homework amount mean value of the target students with the corresponding class homework amount mean value through a second preset algorithm, so that the corresponding first comparison result is used as an homework amount analysis index;

wherein the second preset algorithm comprises:

N-hteNum＝{H ₁ ,H ₂ ,……,H _N }

wherein N-hteNum is the number of times of submitting the operation;

N-hteSubExerciseNum＝{E ₁ ,E ₂ ,……,E _M }

wherein N-hteStubExerciseNum is the number of the operation problems;

wherein, T-stuAugAnsweTime is the actual answer time set of each job, For the actual answer of the 1 st job>For the actual answer of the 2 nd job +.>The actual answer time of the nth operation is the actual answer time;

wherein,time for the work aging transverse contrast value ^student For the actual answer average duration and Time of the current statistical period ^class The average time length average value of actual answers of class students in the current statistical period is calculated;

calculating an operation score stability analysis value of the target student based on the operation average correct rate and the class average correct rate of the target student through a third preset algorithm;

the matching result of the operation average correct rate and the class average correct rate and the operation score stability analysis value are used as the operation correct rate analysis index;

Wherein the third preset algorithm comprises:

wherein G (S_R) is a job fluctuation level characterization value, beta ₀ 、β ₁ 、β ₂ Is a preset parameter;

the determining, by the target student, each dimension reference index in the statistical period specifically includes:

determining the accuracy of the teaching aid/test paper in the operation response information through a fourth preset algorithm; and comparing the teaching aid/test paper accuracy rate, the operation average accuracy rate and the class average accuracy rate, and using the corresponding second comparison result as the analysis index of each teaching aid/test paper;

according to the question serial numbers of the teaching aid/test paper in the operation response information, the question accuracy rates, the question class accuracy rates and the grade question accuracy rates are compared in sequence through the fourth preset algorithm, and corresponding third comparison results are used as error question analysis indexes;

determining the merits and merits of the target students as the merits and merits question analysis index through the fourth preset algorithm; the merits include: master better questions, error-prone questions and important attention questions; the better mastered questions meet the condition that the questions are correctly answered and the corresponding grade questions are smaller than a first preset threshold; the error-prone questions meet that the corresponding grade question accuracy rate is smaller than a second preset threshold; the important attention questions meet the question answering errors, and the corresponding grade question accuracy rate is larger than a third preset threshold;

The fourth preset algorithm comprises the following steps:

grad, egs, represents the average correctness of individual questions in the teaching aid/test paper;

based on each dimension reference index, each analysis and evaluation text is matched, and the method specifically comprises the following steps:

determining a student work accuracy rate sub-model and a work stability sub-model according to the first preset algorithm and the third preset algorithm so as to match analysis evaluation texts corresponding to the student work accuracy rate dimension and the work stability dimension;

determining an operation timeliness management sub-model and an operation accuracy sub-model according to the first preset algorithm and the second preset algorithm so as to match analysis evaluation texts corresponding to the operation timeliness management dimension and the class operation accuracy dimension;

matching the corresponding analysis evaluation text of the target student according to the fourth preset algorithm, the teaching aid/test paper analysis indexes and the merit and inferiority question analysis indexes;

according to the dimension reference indexes, matching preset dialects corresponding to the dimensions; the speaking operation is analysis and evaluation text of preset emotion colors;

The method comprises the steps of comparing historical job information with job response information to determine whether the job response information is analyzable job response information according to a comparison result of the historical job information and the job response information, and specifically comprises the following steps:

calculating cosine similarity of the historical operation information and the operation response information;

under the condition that the cosine similarity is larger than a preset value, determining the job response information as analyzable job response information;

generating alarm information and sending the alarm information to a corresponding management terminal under the condition that the cosine similarity is smaller than a preset value; the alarm information is used for reminding a management terminal personnel that the homework response information of the target student is inconsistent with the historical homework information, and the management terminal personnel selects whether to continue homework analysis on the target student or update the selection range of the historical homework information of the target student;

according to each analysis and evaluation text and the evaluation word set corresponding to the target student, generating an operation analysis report with a corresponding statistical period, wherein the operation analysis report specifically comprises the following steps:

matching corresponding language information sets according to the personal information of the target students; the personal information includes: age, sex, hobbies; the mood information set comprises a plurality of emotion mood words;

Matching each emotion word of the mood information set with each language color of the analysis evaluation text to remove emotion words with unmatched language colors from the mood information set so as to obtain the evaluation mood word set;

training a preset decoder through the evaluation language-gas word set; the decoder is used for outputting a text corresponding to the evaluation language-gas word set;

inputting each analysis and evaluation text into a preset encoder to obtain corresponding encoding vectors; and inputting the coded vector into the trained decoder to generate the job analysis report.

2. The method of claim 1, wherein after generating job analysis reports for respective statistical periods, the method further comprises:

the homework analysis report is sent to a user terminal, so that the homework analysis report corresponding to the target student is displayed at the user terminal; the usage object of the user terminal at least comprises one or more of the following roles: parents, students, teachers.

3. The method according to claim 2, wherein the job analysis report is sent to a user terminal, so as to display the job analysis report corresponding to the target student at the user terminal, and specifically comprises:

Determining a user terminal accessing a link corresponding to the homework analysis report of the target student and an access frequency thereof;

4. A job analysis device, the device comprising:

the acquisition module acquires historical homework information of the target students and current homework response information of the target students; the history job information comprises the history job answering time of each question type and the history job short text; the history homework short text is a text of the target student for expressing homework question answers;

the comparison module is used for comparing the historical job information with the job response information so as to determine whether the job response information is analyzable job response information according to the comparison result of the historical job information and the job response information;

The input module is used for inputting the homework response information into a preset comprehensive evaluation model so as to sequentially match a plurality of analysis evaluation texts of the target students in a preset evaluation dimension set; wherein the set of evaluation dimensions comprises: student homework accuracy dimension, homework stability dimension, homework time efficiency management dimension and class homework accuracy dimension;

the generation module is used for generating an operation analysis report with corresponding statistical period according to each analysis and evaluation text and the evaluation word set corresponding to the target student; the evaluation word set comprises emotion words used for matching and expressing the operation analysis report;

the input module is specifically used for:

Wherein, the first preset algorithm comprises:

the input module is specifically used for:

wherein the second preset algorithm comprises:

N-hteNum＝{H ₁ ,H ₂ ,……,H _N }

wherein N-hteNum is the number of times of submitting the operation;

N-hteSubExerciseNum＝{E ₁ ,E ₂ ,……,E _M }

wherein N-hteStubExerciseNum is the number of the operation problems;

wherein N-wrungHtesubsuccinium is the number of wrong questions, and M is the number of questions of wrong questions and is less than or equal to M;

wherein, T-teacherSuggestTime is a set of recommended duration of a single operation of a teacher, To be at H ₁ The recommended duration of the secondary job;

wherein,for the actual answer of each operation, aggregate, < >>For the actual answer of the 1 st job>For the actual answer of the 2 nd job +.>The actual answer time of the nth operation is the actual answer time;

the input module is specifically used for:

Wherein the third preset algorithm comprises:

wherein G (S_R) isJob fluctuation level characterization value, beta ₀ 、β ₁ 、β ₂ Is a preset parameter;

the input module is specifically used for:

The fourth preset algorithm comprises the following steps:

the input module is specifically used for:

wherein, the comparison module is specifically used for:

the generating module is specifically configured to: