CN110046667B - Teaching evaluation method based on deep neural network learning scoring data pair - Google Patents
Teaching evaluation method based on deep neural network learning scoring data pair Download PDFInfo
- Publication number
- CN110046667B CN110046667B CN201910319686.3A CN201910319686A CN110046667B CN 110046667 B CN110046667 B CN 110046667B CN 201910319686 A CN201910319686 A CN 201910319686A CN 110046667 B CN110046667 B CN 110046667B
- Authority
- CN
- China
- Prior art keywords
- course
- courses
- teaching evaluation
- student
- teaching
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000011156 evaluation Methods 0.000 title claims abstract description 144
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 41
- 238000000034 method Methods 0.000 claims abstract description 42
- 239000013604 expression vector Substances 0.000 claims abstract description 29
- 239000013598 vector Substances 0.000 claims description 14
- 230000006870 function Effects 0.000 claims description 10
- 230000008439 repair process Effects 0.000 claims description 3
- 230000008569 process Effects 0.000 description 11
- 239000011159 matrix material Substances 0.000 description 7
- 230000000694 effects Effects 0.000 description 5
- 230000004913 activation Effects 0.000 description 4
- 238000012549 training Methods 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000010835 comparative analysis Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000012854 evaluation process Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000036651 mood Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/253—Fusion techniques of extracted features
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/20—Education
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Tourism & Hospitality (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- General Health & Medical Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Strategic Management (AREA)
- Biophysics (AREA)
- Educational Technology (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Educational Administration (AREA)
- General Business, Economics & Management (AREA)
- Biomedical Technology (AREA)
- Economics (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Electrically Operated Instructional Devices (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention provides a teaching evaluation method based on deep neural network learning scoring data pair, which comprises the following steps: the method comprises the steps of constructing a student course selection graph by taking courses as nodes and student course selection as sides, dividing the courses into a plurality of course groups by a graph clustering method, wherein the courses in the same course group have similar course selection student groups, and the courses in different course groups have different course selection student groups; in each class group, generating a teaching evaluation pair according to a teaching evaluation database of the students for the courses, wherein the teaching evaluation pair is a teaching evaluation result of one student for two courses; learning a teaching evaluation pair through a deep neural network to obtain a teaching evaluation expression vector of each course; clustering the teaching evaluation expression vectors of each course, dividing a course group into a plurality of course groups, and then obtaining the final course evaluation result by counting teaching evaluation pairs among different course groups. The invention can obtain a fair and fair teaching evaluation result.
Description
Technical Field
The invention relates to the technical field of teaching evaluation methods, in particular to a teaching evaluation method based on deep neural network learning scoring data pairs.
Background
According to the spirit of the reform and the outline of development and planning of long-term education in China (2010-2020) and the opinion of education department on the evaluation work of the teaching of the department of general higher schools, the teaching quality of colleges and universities in the department is improved, the content construction of the department is highlighted, the guarantee of the teaching quality is improved, and the method is an important aspect for continuously improving the culture quality of talents. The course teaching quality guarantee is one of the most basic and most core contents of the colleges and universities education teaching quality guarantee and is also an important content of higher education assessment.
The teaching evaluation method can greatly promote the enthusiasm of teachers in teaching and improve the teaching effect. However, at present, most of domestic and foreign teaching effect assessment depends on evaluation of students on current courses, and then simple analysis and mining are performed on the basis. These methods can be largely classified into two categories. The first category is teaching evaluation based on statistical analysis of students on course evaluation data, and the method ranks the course evaluation by calculating statistics such as average score, variance and the like. The method has single information source and high requirement on the reliability of the original scoring data. However, the real data is often noisy, the data is not standard, for example, the score given to the lesson is influenced by factors such as the examination score, the mood of students when scoring, and the strictness of teachers to classroom management, and in addition, the scoring standards of each student are not uniform. Therefore, simply using the student's score for the course to evaluate the teaching effect of the course often fails to objectively and fairly reflect the teaching level and quality of the course.
The second method is to perform teaching evaluation by a data mining method, for example, an association rule method is used to mine an implicit association relationship between the evaluation result of a student and basic information of a teacher in the class such as gender, academic history, age, and title, but this method cannot effectively integrate different types of features such as student scores and student score examination information to perform comprehensive evaluation.
In recent years, with the rapid development of information technology and the increasing of informatization investment of colleges and universities, big data has already entered the campus of each college, how to really and effectively utilize the big data generated in school teaching and management to perform education and teaching evaluation, and ensuring the objectivity and justice of the teaching evaluation becomes an important problem in teaching evaluation research. Two major challenges exist in current course teaching evaluation: (1) different students have different psychological evaluation standards, so that the scores of different students on the same course are not comparable, therefore, the teaching quality between different courses is compared by simply calculating the average score of all students on the course evaluation, and the result is often unsatisfactory; (2) various types of data can be generated in the teaching process, such as student scores, grades given by students to courses, course types and the like, and how to comprehensively consider various types of data characteristics to evaluate and grade course teaching is a difficult problem.
Disclosure of Invention
Aiming at the defects, the invention provides a teaching evaluation method based on deep neural network learning scoring data pair, so as to overcome the problems that different students have different psychological scoring standards and the characteristics of heterogeneous data are fused and evaluated in the scoring process in the existing method, and thus, the teaching evaluation is more objective and fair.
A teaching evaluation method based on deep neural network learning scoring data pairs comprises the following steps:
the method comprises the steps of constructing a student course selection graph by taking courses as nodes and student course selection as sides, dividing the courses into a plurality of course groups by a graph clustering method, wherein the courses in the same course group have similar course selection student groups, and the courses in different course groups have different course selection student groups;
in each class group, generating a teaching evaluation pair according to a teaching evaluation database of the students for the courses, wherein the teaching evaluation pair is a teaching evaluation result of one student for two courses;
learning a teaching evaluation pair through a deep neural network to obtain a teaching evaluation expression vector of each course;
clustering the teaching evaluation expression vectors of each course, dividing a course group into a plurality of course groups, and then obtaining the final course evaluation result by counting teaching evaluation pairs among different course groups.
The teaching evaluation method based on the deep neural network learning scoring data pair comprises the following steps of constructing a student course selection graph by taking courses as nodes and student course selection as edges, and dividing the courses into several course groups by a graph clustering method:
the courses are taken as nodes, if w students select and repair two courses simultaneously, an edge is arranged between the two courses, the weight of the edge is w, all the courses form a weighted graph, and then all the courses are divided into several courses by utilizing a graph clustering method.
The teaching evaluation method based on the deep neural network learning scoring data pair includes the following steps of:
in each class group, generating a scoring pair according to the scoring of each student on the curriculum taken by each student<s i ,s j >That is, a student scores class i higher than class j, and if a student chooses n classes, n × (n-1)/2 teaching score pairs can be generated for the student.
The teaching evaluation method based on the deep neural network learning scoring data pair includes the following steps of:
And respectively inputting the characteristics of the two courses in the teaching scoring pair into two deep neural networks according to the characteristics of the students, and learning and predicting the teaching scoring of the students on the two courses to learn the teaching evaluation expression vectors of the two courses.
The teaching evaluation method based on the deep neural network learning scoring data pair includes the steps of clustering teaching evaluation expression vectors of each course, dividing a course group into a plurality of course groups, and obtaining a final course evaluation result by counting teaching evaluation pairs among different course groups, wherein the teaching evaluation expression vectors include:
and clustering the course teaching evaluation expression vectors in the course groups aiming at each course group to obtain k course groups, finally dividing the course teaching evaluation into k levels, counting course teaching evaluation pairs among different course groups, and determining the teaching evaluation level of each course group.
The teaching evaluation method based on the deep neural network learning scoring data pair is characterized in that the method further comprises the following steps:
firstly, inputting the characteristics of students into a multilayer neural network, and learning to obtain a student characteristic code; respectively inputting the characteristics of two courses in the course teaching evaluation pair into a multilayer neural network with the same parameters to respectively obtain two course teaching evaluation expression vectors; then, respectively connecting the student characteristic vectors with the two course teaching evaluation expression vectors to form two (student characteristic-course teaching evaluation expression) vectors, respectively inputting a multilayer neural network with the same parameters, and predicting the scores of the students on the two courses; and finally, inputting the prediction score and the real score of the two courses into a loss function to perform learning course teaching evaluation expression vectors.
According to the teaching evaluation method based on the deep neural network learning score data pair, courses in the same course evaluation group in the obtained course evaluation results have the same teaching evaluation result.
The teaching evaluation method based on the deep neural network learning scoring data pair utilizes the deep neural network to learn the comparative evaluation of the same student to different courses, so that the teaching evaluation vector representation of the courses is learned, and then the teaching evaluation grade of the courses is obtained by clustering the course teaching evaluation vectors.
Compared with the prior art, the method comprehensively evaluates the teaching effect grade by combining the learned teaching evaluation data pair and various types of data generated in the teaching process by utilizing the deep learning method, and has more objective and fair results compared with the traditional method, and the specific advantages are that:
(1) as the same student compares the teaching effects of two different courses, the student can more easily give the answer, therefore, the invention obtains the course teaching evaluation expression vector by learning the comparative evaluation of the same student on different courses, and the result is more objective.
(2) The invention designs a new deep neural network architecture for learning, teaching and evaluating data pairs, and in the architecture, various characteristics can be fused: such as student characteristics, student examination scores, course teaching scores, course types and the like, and comprehensive teaching evaluation is performed.
Drawings
Fig. 1 is a schematic flowchart of a teaching evaluation method based on deep neural network learning score data pair according to an embodiment of the present invention;
FIG. 2 is an exemplary diagram of student selected course relationship;
FIG. 3 is a schematic flow chart for generating course evaluation data pairs;
FIG. 4 is a diagram of a deep neural network framework for course evaluation vector representation learning;
FIG. 5 is a schematic view of a course teaching ranking process;
FIG. 6 is a flow chart illustrating the statistical process of evaluation data pairs between lesson groups.
Detailed Description
To facilitate an understanding of the invention, the invention will now be described more fully with reference to the accompanying drawings. Several embodiments of the invention are presented in the drawings. This invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used herein, the term "and/or" includes any and all combinations of one or more of the associated listed items.
Referring to fig. 1, an embodiment of the present invention provides a teaching evaluation method based on a deep neural network learning score data pair, including the following steps S101 to S104:
s101, constructing a student course selection graph by taking courses as nodes and student course selection as sides, and dividing the courses into a plurality of course groups by a graph clustering method, wherein the courses in the same course group have similar course selection student groups, and the courses in different course groups have different course selection student groups;
the course group generation takes courses as nodes, if w students select and repair two courses simultaneously, an edge is arranged between the two courses, the weight of the edge is w, all the courses form a weighted graph, and then all the courses are divided into several course groups by utilizing a graph clustering method.
The specific implementation process is as follows: the purpose of the course group generation is to divide all courses in the whole school into several parts, each part is a course group, and then the teaching quality comparison is carried out in the course group, because the standards of the teaching evaluation of two completely different types of courses may be different, for example, the teaching quality comparison of a literature course and a computer course together may be difficult and meaningless. In order to generate a course group, the invention first constructs a weighted undirected graph G { V, E, W }, wherein V represents a node set, E represents an edge set, and W represents a weight of E, using courses as nodes and the number of students who have selected two courses at the same time as weights, as shown in fig. 2.
In the course group generation implementation, a hierarchy-based graph clustering algorithm is adopted, namely each course is regarded as an independent cluster, and then the following two steps are repeated: (1) computing any two clusters C i ，C j Similarity between them D ij Degree of similarity d ij The average of the weights between the nodes in the two clusters is used as the similarity of the two clusters, namely:
wherein w pq As the weight of the nodes p, q, | C i |、|C j Each is a cluster C i ，C j The number of middle nodes; (2) merging two clusters with the maximum similarity in the existing clusters; repeating the two steps until the number of the clusters reaches a specified threshold value; finally, according to the formula
And combining the independent nodes and the clusters with less nodes generated in the previous step into the cluster with the maximum similarity to obtain the class group required by the invention.
S102, in each class group, generating a teaching evaluation pair according to a teaching evaluation database of the students for the courses, wherein the teaching evaluation pair is a teaching evaluation result of one student for two courses;
wherein, generating scoring data pairs, and generating scoring pairs according to the scoring of the curriculum taken by each student in each class group<s i ,s j >That is, a student scores class i higher than class j, and if a student chooses n classes, n × (n-1)/2 teaching score pairs can be generated for the student.
The specific implementation process is as follows: the score data pairs are generated for the next step of training the deep neural network to obtain a score representation vector for each lesson. One scoring data pair is a training sample of the neural network. The process of generating the scoring data pair is as shown in fig. 3, firstly reading a course taken by a student from a database, forming a course pair by any two courses, if the two courses belong to the same group, the course pair belongs to an effective scoring data pair, storing the information of the scoring data pair in a corresponding storage position, and marking the group to which the grading data pair belongs. The stored grading data correspondingly comprises student characteristic information, characteristic information of the two courses and grading of the two courses by the students. In the specific implementation of the invention, the student characteristic information includes the class corresponding to the student, the grade of the course taken by the student, and the like; the course characteristics comprise the type of the course, the assessment mode, the college to which the course belongs and the grade of all selected students on the course.
S103, learning a teaching evaluation pair through a deep neural network to obtain a teaching evaluation expression vector of each course;
The course teaching evaluation represents vector learning, two deep neural networks are respectively input according to characteristics of students and characteristics of two courses in a teaching scoring pair, and the teaching evaluation representing vectors of the two courses are learned through learning and predicting teaching scoring of the students on the two courses.
The specific implementation process is as follows: the invention designs a new neural network architecture to learn the teaching evaluation expression vector of the course. The framework obtains teaching evaluation representation vectors for each course by learning teaching evaluation pairs. The new neural network architecture is shown in fig. 4, in which student characteristics, as well as two class characteristics, are used as inputs. The student characteristics are input into the student characteristic coding network to obtain student characteristic codes, and the two courses characteristics are respectively input into a course characteristic coding network with the same parameters to respectively obtain the codes of the two courses. The student characteristic coding network and the course characteristic coding network are respectively 64 x 32 network and 128 x 32 network, a relu function is used as an activation function, and a batch normalization layer is added before the activation function. After the student codes and the codes of the two courses are obtained, the student codes are respectively connected with the two course codes to form two student-course code vectors, and the two student-course code vectors are input into a scoring network to obtain the scoring of the course by a student. The scoring network is a three-layer network of 64 x 32 x 1, and the relu function is also used as an activation function, and a batch normalization layer is added before the activation function. Finally, subtracting the two course scores and inputting the two course scores and the real scores into a loss function. In the present invention, a squared error loss function is used, i.e. 
Where n is the number of course scoring data pairs, g prd Predict a score gap, g, for the network lb True square difference. And then training the whole network by utilizing back propagation to obtain network parameters.
And S104, clustering the teaching evaluation expression vectors of each course, dividing a course group into a plurality of course groups, and then obtaining the final course evaluation result by counting the teaching evaluation pairs among different course groups.
And in the obtained course evaluation results, courses in the same course evaluation group have the same teaching evaluation result. And (3) course teaching grading, namely clustering course teaching evaluation expression vectors in the course groups aiming at each course group to obtain k course groups, finally dividing the course teaching evaluation into k levels, counting course teaching evaluation pairs among different course groups, and determining the teaching evaluation level of each course group.
The specific implementation process is as follows: after obtaining the teaching evaluation expression vectors of all courses, the invention clusters each course group according to the preset number k of evaluation grades, then counts the course score pairs among different groups, and determines the evaluation grade of each course group, as shown in fig. 5. In the embodiment, the teaching evaluation expression vectors of the courses are clustered by adopting a traditional k-means algorithm. After k lesson groups are obtained through clustering for each lesson group, the k groups need to be ranked. In the implementation process, the invention firstly applies a k × k array matrix S with 0 initially for storing the statistical result of course scoring pairs. The value in the ith row and jth column of the array matrix S represents the number of scoring data pairs for which the course score in the ith course group is higher than the course score in the jth course group. In the array matrix S, the diagonal elements are 0. Then, the calculation of the value of the array matrix S is started, and the flow is shown in fig. 6; for each score pair in a class group, firstly judging whether two courses belong to the same class group, and if the two courses belong to the same class group or have the same score, discarding the two courses; if not, assume that two courses i, j are from the course group C i 、C j If the grade of the course i is higher than that of the course j, the value of the ith row and the jth column of the data matrix S is increased by 1, and if the grade of the course j is higher than that of the course i, the value of the jth row and the ith column of the data matrix S is increased by 1 until all grade data pairs in the course group are counted.
After the statistical result matrix S is obtained, the rank is evaluated by calculating the sum of each row. The lesson group with the highest sum corresponding to the row has the highest rating and represents that the most students consider this lesson group to be better teaching quality than the other lesson groups. The lesson group having the smallest sum corresponds to the lowest rating level representing fewer students considering the teaching quality of the group to be better than other lesson groups.
The teaching evaluation method based on the deep neural network learning scoring data pair comprises the steps of firstly constructing a course selection relation graph of students, and dividing the course selection relation graph through a graph clustering method to obtain course groups. Then, aiming at each class group, producing a teaching scoring data pair according to a student information base, a teaching scoring base and a course information base; each teaching scoring data pair comprises student information, two courses information and the student's score for the two courses. Then, the invention designs a new deep neural network architecture to learn the teaching evaluation expression vector of each course. Finally, the courses in the course group are divided into k groups by clustering course teaching evaluation expression vectors, and the grade of each course group is determined by counting teaching evaluation data pairs between the course groups. The teaching evaluation method based on the deep neural network learning scoring data pair overcomes the result deviation caused by different students having different psychological evaluation standards in the traditional method, and simultaneously considers the influence of the examination score, the student characteristics and the course characteristics of the students on the teaching evaluation in the evaluation process. Therefore, compared with the traditional method, the method can obtain a more objective and fair teaching evaluation result.
The above-mentioned embodiments only express one embodiment of the present invention, and the description thereof is more specific and detailed, but not construed as limiting the scope of the present invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the inventive concept, which falls within the scope of the present invention. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (3)
1. A teaching evaluation method based on deep neural network learning scoring data pairs is characterized by comprising the following steps:
the method comprises the steps of constructing a student course selection graph by taking courses as nodes and student course selection as sides, dividing the courses into a plurality of course groups by a graph clustering method, wherein the courses in the same course group have similar course selection student groups, and the courses in different course groups have different course selection student groups;
in each class group, generating a teaching evaluation pair according to a teaching evaluation database of the students for the courses, wherein the teaching evaluation pair is a teaching evaluation result of one student for two courses;
learning a teaching evaluation pair through a deep neural network to obtain a teaching evaluation expression vector of each course;
Clustering teaching evaluation expression vectors of each course, dividing a course group into a plurality of course groups, and then obtaining a final course evaluation result by counting teaching evaluation pairs among different course groups;
the steps of constructing a student course selection graph by taking courses as nodes and student course selection as sides and then dividing the courses into several course groups by a graph clustering method specifically comprise the following steps:
taking courses as nodes, if w students select and repair two courses simultaneously, an edge is arranged between the two courses, the weight of the edge is w, thus all the courses form a weighted graph, and then all the courses are divided into several courses by utilizing a graph clustering method;
in each class group, the step of generating the teaching evaluation pair according to the teaching evaluation database of the student to the course specifically comprises the following steps:
in each class group, generating teaching evaluation pairs according to the grading of each student on the curriculum repaired by each student<s i ,s j >If a student chooses n courses, n x (n-1)/2 teaching evaluation pairs can be generated for the student;
the step of learning the teaching evaluation pair through the deep neural network to obtain the teaching evaluation expression vector of each course specifically comprises the following steps:
Respectively inputting two deep neural networks according to the characteristics of the student and the characteristics of two courses in the teaching evaluation pair, and learning teaching evaluation expression vectors of the two courses by predicting the teaching scores of the student on the two courses;
the method further comprises the following steps:
firstly, inputting the characteristics of students into a multilayer neural network, and learning to obtain a student characteristic code; respectively inputting the characteristics of two courses in the course teaching evaluation pair into a multilayer neural network with the same parameters to respectively obtain two course teaching evaluation expression vectors; then, the student characteristic vectors are respectively connected with the two course teaching evaluation expression vectors to form two (student characteristic-course teaching evaluation expression) vectors, and the two vectors are respectively input into a multilayer neural network with the same parameters to predict the scores of the students on the two courses; and finally, inputting the prediction score and the real score of the two courses into a loss function to perform learning course teaching evaluation expression vectors.
2. The method as claimed in claim 1, wherein the step of clustering the teaching evaluation expression vectors of each course, dividing a course group into a plurality of course groups, and then obtaining the final course evaluation result by counting the teaching evaluation pairs between different course groups specifically comprises:
And clustering the course teaching evaluation expression vectors in the course groups aiming at each course group to obtain k course groups, finally dividing the course teaching evaluation into k levels, counting course teaching evaluation pairs among different course groups, and determining the teaching evaluation level of each course group.
3. The teaching evaluation method based on deep neural network learning score data pair as claimed in any one of claims 1 to 2, wherein the courses in the same course evaluation group have the same teaching evaluation result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910319686.3A CN110046667B (en) | 2019-04-19 | 2019-04-19 | Teaching evaluation method based on deep neural network learning scoring data pair |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910319686.3A CN110046667B (en) | 2019-04-19 | 2019-04-19 | Teaching evaluation method based on deep neural network learning scoring data pair |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110046667A CN110046667A (en) | 2019-07-23 |
| CN110046667B true CN110046667B (en) | 2022-08-12 |
Family
ID=67278153
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910319686.3A Active CN110046667B (en) | 2019-04-19 | 2019-04-19 | Teaching evaluation method based on deep neural network learning scoring data pair |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110046667B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110990583B (en) * | 2019-11-25 | 2022-05-10 | 中山大学 | Course map construction method based on network embedding |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2001080153A1 (en) * | 2000-04-15 | 2001-10-25 | Kim Sang Su | Method and system for network-based education and training capable of evaluating learner"s faithfulness and computer-readable medium recording its program |
| CN106023015A (en) * | 2016-05-18 | 2016-10-12 | 腾讯科技（深圳）有限公司 | Course learning path recommending method and device |
| CN106203811A (en) * | 2016-07-05 | 2016-12-07 | 上海电力学院 | A kind of Evaluation Method of Teaching Quality |
| CN107180568A (en) * | 2017-06-30 | 2017-09-19 | 楼志刚 | A kind of Kernel-based methods control and the interactive learning platform and method of evaluation of result |
| CN107590760A (en) * | 2017-11-01 | 2018-01-16 | 长春理工大学 | A kind of Teaching quality evaluation system based on big data |
| CN108256102A (en) * | 2018-02-01 | 2018-07-06 | 厦门大学嘉庚学院 | A kind of Independent College Studentss based on cluster comment religion data analysing method |
| CN109255028A (en) * | 2018-08-28 | 2019-01-22 | 西安交通大学 | Quality of instruction integrated evaluating method based on teaching evaluation data reliability |
| CN109360458A (en) * | 2018-10-25 | 2019-02-19 | 重庆鲁班机器人技术研究院有限公司 | Interest assistant teaching method, device and robot |
| CN109597937A (en) * | 2018-12-03 | 2019-04-09 | 华中师范大学 | Network courses recommended method and device |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9152884B2 (en) * | 2012-06-05 | 2015-10-06 | Drvision Technologies Llc | Teachable pattern scoring method |
-
2019
- 2019-04-19 CN CN201910319686.3A patent/CN110046667B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2001080153A1 (en) * | 2000-04-15 | 2001-10-25 | Kim Sang Su | Method and system for network-based education and training capable of evaluating learner"s faithfulness and computer-readable medium recording its program |
| CN106023015A (en) * | 2016-05-18 | 2016-10-12 | 腾讯科技（深圳）有限公司 | Course learning path recommending method and device |
| CN106203811A (en) * | 2016-07-05 | 2016-12-07 | 上海电力学院 | A kind of Evaluation Method of Teaching Quality |
| CN107180568A (en) * | 2017-06-30 | 2017-09-19 | 楼志刚 | A kind of Kernel-based methods control and the interactive learning platform and method of evaluation of result |
| CN107590760A (en) * | 2017-11-01 | 2018-01-16 | 长春理工大学 | A kind of Teaching quality evaluation system based on big data |
| CN108256102A (en) * | 2018-02-01 | 2018-07-06 | 厦门大学嘉庚学院 | A kind of Independent College Studentss based on cluster comment religion data analysing method |
| CN109255028A (en) * | 2018-08-28 | 2019-01-22 | 西安交通大学 | Quality of instruction integrated evaluating method based on teaching evaluation data reliability |
| CN109360458A (en) * | 2018-10-25 | 2019-02-19 | 重庆鲁班机器人技术研究院有限公司 | Interest assistant teaching method, device and robot |
| CN109597937A (en) * | 2018-12-03 | 2019-04-09 | 华中师范大学 | Network courses recommended method and device |
Non-Patent Citations (8)
| Title |
|---|
| "A Kind of Analytic Hierarchy Process and Neural Network Integrated Evaluation Method of Teaching";Yang H等;《2016 International Conference on Humanity，Education and Social Science》;20161231;第231-235页 * |
| "cluster analysis on the teaching evaluation data from college students";Pingping Gu等;《2018 13th International Conference on Computer Science & Education》;20180930;第839-844页 * |
| "Teaching Evaluation and Diagnosis Based on BP Neural Network";Hu Zhijun等;《2009 International Conference on Management and Service Science》;20091231;第1-3页 * |
| "基于混合高斯模型融合历史评价信息的高校教学评价方法";黄晓辉等;《亚太教育》;20161231(第36期);第153页及第118页 * |
| "基于秩和检验及非参数Bootstrap法的多指标统计综合评价模型";晏榆洋;《四川文理学院学报》;20180930;第28卷(第5期);第12-15页 * |
| "大数据环境下基于多维信息融合的高校教学评价";黄晓辉等;《亚太教育》;20161231(第35期);第199-200页 * |
| "大数据视角下高校网络教学平台应用的实践与反思";王佳利;《中国现代教育装备》;20151130(第229期);第13-16页 * |
| "课程建设质量评价体系的神经网络模型";施蕙;《江苏技术师范学院学报》;20040630;第10卷(第2期);第53-56页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110046667A (en) | 2019-07-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wilde et al. | How close is close enough? Evaluating propensity score matching using data from a class size reduction experiment | |
| CN113344053B (en) | Knowledge tracking method based on examination question different composition representation and learner embedding | |
| Fong et al. | Applying a hybrid model of neural network and decision tree classifier for predicting university admission | |
| Hamim et al. | Student profile modeling using boosting algorithms | |
| CN112380429A (en) | Exercise recommendation method and device | |
| CN110046667B (en) | Teaching evaluation method based on deep neural network learning scoring data pair | |
| Zhang et al. | Research and application of grade prediction model based on decision tree algorithm | |
| CN115617960A (en) | Post recommendation method and device | |
| Gray et al. | An investigation of psychometric measures for modelling academic performance in tertiary education | |
| Adeyemo et al. | Mining students’ academic performance using decision tree algorithms | |
| Binh et al. | Student ability estimation based on IRT | |
| Nuankaew et al. | Self-Regulated Learning Styles in Hybrid Learning Using Educational Data Mining Analysis | |
| Gu et al. | Cluster analysis on the teaching evaluation data from college students | |
| Liu et al. | Investigating the use of fuzzy logical data for evaluating the teaching quality of college teachers | |
| Valsamidis et al. | A combined algorithm for LMS usage assessment | |
| CN112200474A (en) | Teaching quality evaluation method, terminal device and computer readable storage medium | |
| Irfiani et al. | Predicting grade promotion using decision tree and Naïve Bayes classification algorithms | |
| Gu | [Retracted] Evaluation of Teaching Quality on IP Environment Driven by Multiple Values Theory Based on Big Data | |
| Dake et al. | Students Grades Predictor using Naïve Bayes Classifier–A Case Study of University of Education, Winneba | |
| Nidhi et al. | Predictive Model for Students' Academic Performance Using Classification and Feature Selection Techniques | |
| Jia | [Retracted] Random Algorithm and Skill Evaluation System Based on the Combing of Construction Mechanism of Higher Vocational Professional Group | |
| Hasan et al. | Make the educational decisions using the analytical hierarchy process AHP in the light of the corona pandemic | |
| Vakhitov et al. | Clustering of social network data by means of Korhonen neural networks | |
| Alblawi et al. | Leveraging Prerequisites Engineering Students for Monitoring and Predicting Using Machine Learning | |
| Dong et al. | Discovering the associations between test items in an examination |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |