An Accurate Knowledge Service Recommendation Method for College Ideological Education Based on Data Portrait Technology

To realize accurate knowledge service recommendation, it is necessary to design a special recommendation algorithm to realize personalized recommendation based on students’ feedback information and behavioral data. It is one of the difficulties to analyze students’ learning habits, hobbies, knowledge level and other personality characteristics, and to extract the landmarks of knowledge service. Therefore, a new method of accurate knowledge service recommendation for ideological and political education in colleges and universities based on data portrait technology is proposed. Using data profiling technology to extract knowledge markers, analyzing individual characteristics such as learning habits, interests, and knowledge levels of students, to extract knowledge markers related to educational content. On this basis, a probability model learning method combined with association rules was used to analyze the extracted knowledge markers, and a feature mining model for precise knowledge services in ideological and political education in universities was established. The educational content was classified to distinguish between high-quality and low-quality content. Timestamps are used to divide the content that users have been interested in in the past, and to obtain the rating vector for each user's item by calculating the rating matrix, in order to identify which content still has high value and relevance over time. The recursive neural network is used to realize the accurate knowledge service recommendation. The experimental results show that the distribution of educational knowledge service data obtained by data mapping is consistent with the actual situation. Moreover, the recommended results of the research method have higher similarity and confidence.