Recommendation Model of Mathematical Expressions Based on Collaborative Filtering

Abstract: The rapid development of the Internet has caused the problem of “information overload”, especially in the field of information retrieval. Since mathematical retrieval as part of information retrieval, this paper proposed a model of interest recommendation for it. First, the public data sets of mathematical expressions were normalized for the mathematical formulas which have the same computing meaning but different parameters. Secondly, the concept of fuzzy set was employed for making the system to know the similarity of users according to the different habits of each user in the multi features fuzzy pattern through integrating these attributes to form a “user-formula” scoring matrix for calculating the closeness degree between users. And the nearest neighbor set of the target user is obtained. Finally, using the fuzzy elements of score and the corresponding neighbors to predict the score of the target user with non- scoring formulas to form the recommendation list of interested formulas through the further filtering of the threshold factor, and realize the personalized recommendation for the mathematical expressions at last. The feasibility and effectiveness of the proposed method are verified by the experiment and data.

Keywords: Mathematical expressions, collaborative filtering, fuzzy pattern with multi features, normalization, personalized recommendation