Intelligent evaluation method for ideological and political education effect
Through cross-platform data crawling and deep learning technology, an evaluation model for ideological and political education effect was constructed, which solved the problem of insufficient evaluation of ideological and political education effect on new media platforms, and achieved the ability to comprehensively evaluate the communication effect of ideological and political education and optimize the communication strategy.
Patent Information
- Application Number
- CN202411542054.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-31
- Publication Date
- 2025-02-28
AI Technical Summary
The existing technology lacks systematic cross-platform research, especially in evaluating the educational achievements, interactivity, and popularity of audience participation in ideological and political education on new media platforms.
Cross-platform data crawling, big data analysis, sentiment analysis and deep learning are used to build an evaluation model for the effect of ideological and political education, and data is captured from multiple new media platforms through online crawlers, and preprocessing and quantitative evaluation are performed to show the communication effect of ideological and political education on different new media platforms.
A comprehensive and dynamic assessment of the effectiveness of ideological and political education on new media platforms has been achieved, helping educators improve communication efficiency, improve students' participation and learning effectiveness, and optimize communication strategies.
Smart Images

Figure 00000000_0000_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of computer data processing in education, and in particular to an intelligent evaluation method for the effect of ideological and political education. Background Art
[0002] With the rapid development of new media technology, traditional ideological and political education methods are facing challenges. New media platforms such as WeChat, Weibo, and Douyin have unique advantages in disseminating information and interaction, which increases the speed and breadth of the dissemination of ideological and political education content. How to use new media and big models to enhance the influence of ideological and political education has become a current research hotspot. However, how to effectively evaluate the acceptance of ideological and political education on these new media platforms and improve the audience's participation and educational effects is still the current research focus.
[0003] At present, domestic and foreign research mainly focuses on the analysis of the impact of new media platforms on ideological and political education, but there is a lack of systematic cross-platform research, especially in the evaluation of educational outcomes, interactivity, and audience participation enthusiasm. Summary of the invention
[0004] The technical problem to be solved by the present invention is to provide an intelligent evaluation method for the effect of ideological and political education, which comprehensively and dynamically evaluates the effect of ideological and political education under the dissemination of new media platforms through cross-platform data capture, big data analysis, sentiment analysis and deep learning.
[0005] The technical solution of the present invention is:
[0006] An intelligent evaluation method for the effect of ideological and political education specifically includes the following steps:
[0007] (1) Use web crawlers to crawl data on the dissemination of ideological and political education content from multiple new media platforms and store it in a local database; the dissemination of ideological and political education content includes published ideological and political education works and user participation and interaction data related to ideological and political education works. Each ideological and political education work is regarded as a comment object;
[0008] (2) Preprocess the data stored in the local database;
[0009] (3) Construct an ideological and political education effect evaluation model, based on multi-dimensional indicators, and use machine learning and natural language processing technology to quantitatively evaluate the ideological and political education effect of pre-processed data;
[0010] (4) After the evaluation of the ideological and political education effect evaluation model is completed, the evaluation results of the data are analyzed, and the communication effects of ideological and political education on different new media platforms are displayed through data visualization tools;
[0011] (5) Based on the quantitative evaluation results of the ideological and political education effectiveness evaluation model, comprehensively judge the degree of dissemination of ideological and political education.
[0012] When the web crawler searches and grabs data on the set ideological and political education works from multiple new media platforms, it first inputs the set ideological and political education works to be searched and crawled, grabs data on the set ideological and political education works that are disseminated on multiple new media platforms, and stores them separately.
[0013] When the web crawler captures data on the dissemination of ideological and political education content, its strategy is: to build a database containing ideological and political education related keywords, filter out content related to the dissemination of ideological and political education through ideological and political education related keywords, use a set time span to obtain the dissemination data of ideological and political education content on different new media platforms in different time periods, and finally save the captured data on the dissemination of ideological and political education content on different new media platforms.
[0014] The data stored in the local database is preprocessed, the preprocessing includes deduplication, deletion of irrelevant content and standardization;
[0015] The deduplication mentioned above is to delete duplicate content and redundant data to ensure the uniqueness of the analyzed data;
[0016] The said deletion of irrelevant content is to filter out the content irrelevant to ideological and political education by screening the keywords related to ideological and political education;
[0017] The standardization process is to standardize the numerical data in the stored data to eliminate the differences in data scales between different new media platforms.
[0018] The specific steps of constructing the ideological and political education effect evaluation model, based on multi-dimensional indicators, using machine learning and natural language processing technology to quantitatively evaluate the ideological and political education effect of pre-processed data are as follows:
[0019] S31, using the user participation interaction data of each evaluation object at the current time t as an evaluation index, and calculating the participation heat score of each evaluation object;
[0020] S32, using machine learning and natural language processing technology to perform sentiment analysis on each evaluation object at the current time t, and calculate the sentiment tendency score of each evaluation object at the current time t;
[0021] S33, calculating the similarity between each comment object and the user comments in the user interaction data at the current time t, as the topic relevance score of the evaluation object at the current time t;
[0022] S34. The participation heat score, sentiment tendency score, and topic relevance score of each evaluation object at the current time t are used as indicators of three dimensions, and weight calculation is performed to obtain a comprehensive score of each evaluation object at the current time t.
[0023] The calculation steps of the participation heat score are as follows:
[0024] S311, the user participation interaction data of each new media platform at the current time t is processed by decimal calibration and normalization, and then the processed user participation interaction data of each evaluation object is filled into the original data matrix A = (a ij ) n×m In, a ij represents the jth indicator of the i-th evaluation object;
[0025] S312, calculate the proportion p of the jth index value of the i-th evaluation object ij , see the following formula (1) for details:
[0026]
[0027] In formula (1), n is the total number of evaluation objects in the original data matrix A, and m is the total number of indicators in the original data matrix A;
[0028] S313. Calculate the entropy value e of the jth index j , see the following formula (2) for details:
[0029]
[0030] S314. Calculate the coefficient of variation g of the jth indicator j , see the following formula (3) for details:
[0031] g j =1-e j ,j=1,2,...,m (3);
[0032] S315. Calculate the weight w of the jth indicator j , see the following formula (4) for details:
[0033]
[0034] S316. Calculate the participation heat score s of the i-th evaluation object i , see the following formula (5) for details:
[0035]
[0036] The calculation steps of the sentiment tendency score are as follows: first, based on the transformer library, the BERT model is used to perform sentiment classification on the user comments of each evaluation object at the current time t, so that each user comment is divided into three types of sentiment labels: positive, negative and neutral. The three types of sentiment labels correspond to corresponding sentiment scores. The sentiment scores of all user comments of each evaluation object are counted and added up, that is, the sentiment tendency score of each evaluation object at the current time t is obtained.
[0037] The steps for calculating the topic relevance score are as follows: using the BERT model to convert the evaluation object and the user comment text data of the evaluation object at the current time t into vectors, calculating the cosine similarity between the evaluation object and each user comment, and after calculating the cosine similarity between each user comment and the evaluation object, taking the average of all cosine similarities calculated for this evaluation object to obtain the average similarity value, and using the average similarity value as the topic relevance score of the evaluation object at the current time t.
[0038] The specific contents of analyzing the evaluation results of the data and displaying the dissemination effect of ideological and political education on different new media platforms through data visualization tools include: (1) calculating the comprehensive scores of ideological and political education works published by the same user on the same new media platform in different time periods, and then taking the different time periods as the horizontal axis and the comprehensive scores of works published in different time periods as the vertical axis, making a line graph and visualizing it; (2) calculating the comprehensive scores of the same ideological and political education works on different new media platforms, and then taking the different new media platforms as the horizontal axis and the comprehensive scores of the same ideological and political education works on different new media platforms as the vertical axis, making a line graph and visualizing it.
[0039] The specific content of comprehensively judging the degree of dissemination of ideological and political education based on the quantitative evaluation results of the ideological and political education effect evaluation model is as follows: setting a threshold value. When the comprehensive score of the evaluation object is not lower than the threshold value, it means that the evaluation object, i.e., the current ideological and political education content, has a good degree of dissemination on this new media platform. When the comprehensive score of the evaluation object is lower than the threshold value, it means that the evaluation object, i.e., the current ideological and political education content, has a low degree of dissemination on this new media platform, reminding users that they need to adjust their dissemination strategies.
[0040] Advantages of the present invention:
[0041] (1) This paper studies the acceptance outcomes of ideological and political education on new media platforms from the perspective of communication, including different forms of modal data, the effects achieved, and the participation enthusiasm absorbed, to promote the effective implementation of ideological and political education.
[0042] (2) The invented and constructed ideological and political education effect evaluation model conducts a quantitative evaluation of the ideological and political education effect based on indicators in three dimensions, helping educators to more efficiently disseminate ideological and political education content on new media platforms, thereby improving students' participation and learning outcomes.
[0043] (3) The present invention can not only effectively evaluate the dissemination effect of ideological and political education on new media platforms, but also intuitively judge the degree of dissemination of ideological and political education through the setting of thresholds, which is convenient for educators to optimize the dissemination strategy on the corresponding multimedia platform. BRIEF DESCRIPTION OF THE DRAWINGS
[0044] Figure 1 It is a flow chart of the present invention.
[0045] Figure 2 It is a statistical line chart of the comprehensive scores of works published by the same user in different time periods on the same new media platform in an embodiment of the present invention.
[0046] Figure 3 It is a statistical line chart of the comprehensive scores of the same ideological and political education work on different new media platforms in an embodiment of the present invention. DETAILED DESCRIPTION
[0047] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0048] See Figure 1 , an intelligent evaluation method for the effect of ideological and political education, specifically including the following steps:
[0049] (1) Use web crawlers to crawl and disseminate data on ideological and political education content from multiple new media platforms and store them in a local database. When crawlers crawl and disseminate data on ideological and political education content, their strategy is to build a database containing keywords related to ideological and political education, filter out content related to ideological and political education dissemination through keywords related to ideological and political education, use a set time span to obtain the dissemination data of ideological and political education content on different new media platforms in different time periods, and finally save the crawled data on the dissemination of ideological and political education content on different new media platforms.
[0050] When the web crawler searches and crawls data on the dissemination of set ideological and political education works from multiple new media platforms, the set ideological and political education works to be searched and crawled are first input, and the data on the dissemination of the set ideological and political education works on multiple new media platforms are crawled and stored separately;
[0051] New media platforms: Weibo, WeChat, Bilibili (Bilibili video sharing website), Zhihu and other mainstream social platforms;
[0052] The content of ideological and political education includes published ideological and political education works and user participation and interaction data related to ideological and political education works. Each ideological and political education work is regarded as a comment object.
[0053] (2) Preprocess the data stored in the local database, including deduplication, deletion of irrelevant content, and standardization;
[0054] Deduplication is to remove duplicate content and redundant data to ensure the uniqueness of the analyzed data;
[0055] Deleting irrelevant content means filtering out content irrelevant to ideological and political education by filtering out keywords related to ideological and political education;
[0056] Standardization is to standardize the numerical data in the stored data to eliminate the differences in data scales between different new media platforms;
[0057] (3) Construct an evaluation model for the effect of ideological and political education. Based on the three-dimensional indicators, use machine learning and natural language processing technology to quantitatively evaluate the effect of ideological and political education on the pre-processed data. The specific steps are as follows:
[0058] S31, using the user participation interaction data of each evaluation object at the current time t as the evaluation index, and calculating the participation heat score of each evaluation object, the calculation steps are:
[0059] S311, the user participation interaction data (reading volume, likes, comments, and favorites) of each new media platform at the current time t is processed by decimal calibration and normalization, and then the processed user participation interaction data of each evaluation object is filled into the original data matrix A = (a ij ) n×m In, a ij represents the jth indicator of the i-th evaluation object; wherein, the original data matrix A is shown in Table 1 below:
[0060]
[0061] S312, calculate the proportion p of the jth index value of the i-th evaluation object ij , see the following formula (1) for details:
[0062]
[0063] In formula (1), n is the total number of evaluation objects in the original data matrix A, and m is the total number of indicators in the original data matrix A;
[0064] S313. Calculate the entropy value e of the jth index j , see the following formula (2) for details:
[0065]
[0066] S314. Calculate the coefficient of variation g of the jth indicator j , see the following formula (3) for details:
[0067] g j =1-e j ,j=1,2,...,m (3);
[0068] S315. Calculate the weight w of the jth indicator j , see the following formula (4) for details:
[0069]
[0070] S316. Calculate the participation heat score s of the i-th evaluation object i , see the following formula (5) for details:
[0071]
[0072] S32. Use machine learning and natural language processing technology to perform sentiment analysis on each evaluation object at the current time t. Specifically, firstly, based on the transformer library, use the BERT model to perform sentiment classification on the user comments of each evaluation object at the current time t, so as to divide each user comment into three types of sentiment labels: positive, negative and neutral. Each of the three types of sentiment labels corresponds to a corresponding sentiment score. The sentiment scores of all user comments on each evaluation object are counted and added up, that is, the sentiment tendency score s of each evaluation object at the current time t is obtained. e ;
[0073] S33, using the BERT model to convert the evaluation object and the user comment text data of the evaluation object at the current time t into vectors, calculate the cosine similarity between the evaluation object and each user comment, and after calculating the cosine similarity between each user comment and the evaluation object, take the average of all cosine similarities calculated for this evaluation object to obtain the average similarity value, and use the average similarity value as the topic relevance score s of the evaluation object at the current time t r ;
[0074] S34: Take the participation heat score, sentiment tendency score, and topic relevance score of each evaluation object at the current time t as indicators of the three dimensions, perform weight calculation, and obtain the comprehensive score s of each evaluation object at the current time t. T , see the following formula (1) for details:
[0075] s T =s i ×0.3+s e ×0.35+s r ×0.3 (1);
[0076] Among them, the constructed ideological and political education effect evaluation model is trained before the prediction and evaluation to obtain a trained ideological and political education effect evaluation model;
[0077] (4) After the evaluation of the ideological and political education effect evaluation model, the evaluation results of the data are analyzed, and the specific contents of the ideological and political education communication effect on different new media platforms are displayed through data visualization tools. (1) The comprehensive scores of ideological and political education works published by the same user in different time periods on the same new media platform are counted, and then different time periods are used as the horizontal axis and the comprehensive scores of works published in different time periods are used as the vertical axis to make a line chart and visualize it (see Figure 2 );
[0078] (2) Count the comprehensive scores of different new media platforms for the same ideological and political education work, then use the different new media platforms as the horizontal axis and the comprehensive scores of different new media platforms for the same ideological and political education work as the vertical axis to make a line graph and visualize it (see Figure 3 );
[0079] According to the needs of users, a sentiment distribution pie chart can be generated for a certain ideological and political education work on a certain new media platform, that is, the sentiment scores of all user comments on the evaluation object are counted, the proportion of three types of sentiment labels for an evaluation object is obtained, and a distribution pie chart is drawn;
[0080] (5) According to the quantitative evaluation results of the ideological and political education effect evaluation model, the degree of dissemination of ideological and political education is comprehensively judged. The specific content is: set a threshold of 0.6. When the comprehensive score of the evaluation object is not less than 0.6, it means that the evaluation object, that is, the current ideological and political education content, has a good degree of dissemination on this new media platform. When the comprehensive score of the evaluation object is less than 0.6, it means that the evaluation object, that is, the current ideological and political education content, has a low degree of dissemination on this new media platform, reminding users that they need to adjust their dissemination strategies.
[0081] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.
Claims
1. An intelligent evaluation method for the effect of ideological and political education, characterized by: The specific steps include: (1) Use web crawlers to crawl data on the dissemination of ideological and political education content from multiple new media platforms and store it in a local database; the dissemination of ideological and political education content includes published ideological and political education works and user participation and interaction data related to ideological and political education works. Each ideological and political education work is regarded as a comment object; (2) Preprocess the data stored in the local database; (3) Construct an ideological and political education effect evaluation model, based on multi-dimensional indicators, and use machine learning and natural language processing technology to quantitatively evaluate the ideological and political education effect of pre-processed data; (4) After the evaluation of the ideological and political education effect evaluation model is completed, the evaluation results of the data are analyzed, and the communication effects of ideological and political education on different new media platforms are displayed through data visualization tools; (5) Based on the quantitative evaluation results of the ideological and political education effectiveness evaluation model, comprehensively judge the degree of dissemination of ideological and political education.
2. According to claim 1, the intelligent evaluation method for the effect of ideological and political education is characterized by: When the web crawler searches, grabs and disseminates data on set ideological and political education works from multiple new media platforms, it first inputs the set ideological and political education works to be searched and crawled, grabs data on the set ideological and political education works disseminated on multiple new media platforms, and stores them separately.
3. According to claim 2, an intelligent evaluation method for the effect of ideological and political education is characterized by: When the web crawler captures data on the dissemination of ideological and political education content, its strategy is: to build a database containing ideological and political education related keywords, filter out content related to the dissemination of ideological and political education through ideological and political education related keywords, use a set time span to obtain the dissemination data of ideological and political education content on different new media platforms in different time periods, and finally save the captured data on the dissemination of ideological and political education content on different new media platforms.
4. According to claim 3, the intelligent evaluation method for the effect of ideological and political education is characterized by: The data stored in the local database is preprocessed, the preprocessing includes deduplication, deletion of irrelevant content and standardization; The deduplication mentioned above is to delete duplicate content and redundant data to ensure the uniqueness of the analyzed data; The said deletion of irrelevant content is to filter out the content irrelevant to ideological and political education by screening the keywords related to ideological and political education; The standardization process is to standardize the numerical data in the stored data to eliminate the differences in data scales between different new media platforms.
5. According to claim 4, an intelligent evaluation method for the effect of ideological and political education is characterized by: The specific steps of constructing the ideological and political education effect evaluation model, based on multi-dimensional indicators, using machine learning and natural language processing technology to quantitatively evaluate the ideological and political education effect of pre-processed data are as follows: S31, using the user participation interaction data of each evaluation object at the current time t as an evaluation index, and calculating the participation heat score of each evaluation object; S32, using machine learning and natural language processing technology to perform sentiment analysis on each evaluation object at the current time t, and calculate the sentiment tendency score of each evaluation object at the current time t; S33, calculating the similarity between each comment object and the user comments in the user interaction data at the current time t, as the topic relevance score of the evaluation object at the current time t; S34. The participation heat score, sentiment tendency score, and topic relevance score of each evaluation object at the current time t are used as indicators of three dimensions, and weight calculation is performed to obtain a comprehensive score of each evaluation object at the current time t.
6. According to claim 5, an intelligent evaluation method for the effect of ideological and political education is characterized by: The calculation steps of the participation heat score are as follows: S311, the user participation interaction data of each new media platform at the current time t is processed by decimal calibration and normalization, and then the processed user participation interaction data of each evaluation object is filled into the original data matrix A = (a ij ) n×m In, a ij represents the jth indicator of the i-th evaluation object; S312, calculate the proportion p of the jth index value of the i-th evaluation object ij , see the following formula (1) for details: In formula (1), n is the total number of evaluation objects in the original data matrix A, and m is the total number of indicators in the original data matrix A; S313. Calculate the entropy value e of the jth index j , see the following formula (2) for details: S314. Calculate the coefficient of variation g of the jth indicator j , see the following formula (3) for details: g j =1-e j ,j=1,2,...,m (3); S315. Calculate the weight w of the jth indicator j , see the following formula (4) for details: S316. Calculate the participation heat score s of the i-th evaluation object i , see the following formula (5) for details:
7. According to claim 5, an intelligent evaluation method for the effect of ideological and political education is characterized by: The calculation steps of the sentiment tendency score are as follows: first, based on the transformer library, the BERT model is used to perform sentiment classification on the user comments of each evaluation object at the current time t, so that each user comment is divided into three types of sentiment labels: positive, negative and neutral. The three types of sentiment labels correspond to corresponding sentiment scores. The sentiment scores of all user comments of each evaluation object are counted and added up, that is, the sentiment tendency score of each evaluation object at the current time t is obtained.
8. According to claim 5, an intelligent evaluation method for the effect of ideological and political education is characterized by: The steps for calculating the topic relevance score are as follows: using the BERT model to convert the evaluation object and the user comment text data of the evaluation object at the current time t into vectors, calculating the cosine similarity between the evaluation object and each user comment, and after calculating the cosine similarity between each user comment and the evaluation object, taking the average of all cosine similarities calculated for this evaluation object to obtain the average similarity value, and using the average similarity value as the topic relevance score of the evaluation object at the current time t.
9. According to claim 5, an intelligent evaluation method for the effect of ideological and political education is characterized by: The specific contents of analyzing the evaluation results of the data and displaying the dissemination effect of ideological and political education on different new media platforms through data visualization tools include: (1) calculating the comprehensive scores of ideological and political education works published by the same user on the same new media platform in different time periods, and then taking the different time periods as the horizontal axis and the comprehensive scores of works published in different time periods as the vertical axis, making a line graph and visualizing it; (2) calculating the comprehensive scores of the same ideological and political education works on different new media platforms, and then taking the different new media platforms as the horizontal axis and the comprehensive scores of the same ideological and political education works on different new media platforms as the vertical axis, making a line graph and visualizing it.
10. According to claim 5, the intelligent evaluation method for the effect of ideological and political education is characterized by: The specific content of comprehensively judging the degree of dissemination of ideological and political education based on the quantitative evaluation results of the ideological and political education effect evaluation model is as follows: setting a threshold value. When the comprehensive score of the evaluation object is not lower than the threshold value, it means that the evaluation object, i.e., the current ideological and political education content, has a good degree of dissemination on this new media platform. When the comprehensive score of the evaluation object is lower than the threshold value, it means that the evaluation object, i.e., the current ideological and political education content, has a low degree of dissemination on this new media platform, reminding users that they need to adjust their dissemination strategies.
Citation Information
Cited By
Education effect evaluation method and system based on artificial intelligence
CN119831186A
