| In recent years,the rapid development of smart education has allowed learners to obtain abundant learning resources and free up more learning time.It has also enabled the platform to accumulate massive amounts of online teaching and learning data.These data are further used to optimize the quality of teaching,thus attracting widespread attention.However,the current online education platform still has the following problems:(1)The quality of online courses is inconsistent,and the evaluation standards urgently need to be standardized;(2)A large number of the same type of courses makes it difficult for learners to distinguish between the pros and cons of different courses,and it is impossible to really recommend personalized teaching services for the learners.Therefore,online education needs to improve the transparency of the courses to allow learners to choose between different courses and find the one appropriate for their needs to provide learners with more concise,efficient,and highquality teaching services.In order to achieve this goal,it is crucial to standardize the evaluation criteria of online courses so that the platform can objectively and comprehensively recommend high-quality courses for learners.Therefore,it is necessary to optimize and innovate online course evaluation methods.In order to achieve this goal,this paper puts forward an indicator system of course portrait,which is a new online course evaluation indicator system for course content.The indicator system includes four first-level indicators and 14 second-level indicators,among which five indicators are implicit features of the course.Firstly,three algorithms have been used to extract the indicator features.The improved Fast Text algorithm is used to classify the course introduction text and extract three implicit features related to the course difficulty.The support vector machine(SVM)algorithm is used to analyze the sentiment of course evaluation,and the implicit features related to course evaluation are extracted.Jaccard similarity is used to carry out fine-grained clustering of similar courses,and the knowledge point coverage of each course is calculated.Furthermore,this paper combined the analytic hierarchy process(AHP)and entropy weight method to determine the weight of all levels of course portrait indicators.Finally,using a linear weighting model,that can reflect the outstanding indicators of course portrait,to calculate the course portrait of each indicator value to achieve the construction of course portraits.Basically,this paper overviews course resources based on the current state of online education and analyzes the advantages and disadvantages of existing education resources.The paper also examines the relevance of current online curricula for learners to create course portraits to help learners identify appropriate course and choose the one for them.Course portraits can allow online platforms to match course characteristics with the right learners to increase relevancy and learner experience.The experimental conception is verified by using accurate data set of an online education platform.The results show that the definition of course portrait,the extraction of course implicit features,and the calculation method of course portrait indicator value have a good rational basis.Compared with the traditional recommendation algorithm,the recommendation model proposed in this paper,which integrates the characteristics of learners and courses,has achieved a significant performance improvement.The research method proposed in this paper has been successfully used to construct a course portrait and recommendation system and a prototype was successfully applied,which provides a powerful reference for standardizing the quality of online courses and improving personalized online teaching. |
PDF Full Text Request