A Personalized Recommendation Approach for Blended English Teaching Database Services Based on Content Retrieval

The blended English teaching database may include non-relational databases and relational databases. The data structure, query method and storage mechanism of each database are different, which increases the difficulty of data integration and analysis, thereby reducing the recommendation validity of personalized recommendations for blended English teaching database services. To this end, a personalized recommendation method for blended English teaching database services based on content retrieval is proposed. Analyze the unique attributes of non-relational and relational databases in the blended English teaching database, accurately define the complex relationships between tables in different databases, and then build a mapping relationship diagram model of the blended English teaching database to achieve the optimization of the database structure graphical conversion. Through cluster analysis of the behavioral data of the service objects of the blended English teaching database, the user’s preference characteristics are captured, and these preference data are efficiently organized in a labeled manner, achieving accurate matching of user preferences. Entropy value vectorization technology is used to efficiently process the service data text in the blended English teaching database. Through content retrieval technology, the similarity between graph data and retrieval data is calculated, and interference factors in the data are effectively eliminated. On this basis, a personalized recommendation model based on content retrieval is constructed to provide users with more accurate and efficient recommendation services. Experimental results show that the proposed method shows significant advantages in improving recommendation effects.