Variance analysis of the project-based similarity and user-based similarity: the above mentioned three similarity formulas are based on the scenario of project similarity, but there is actually a basic difference between project-based similarity and user-based similarity. To be specific, the calculation formula of project-based similarity solves similarity according to the similarity of column vector in the rating matrix, while the calculation formula of user-based similarity solves similarity according to row vector in the rating matrix. Therefore, the above three formula can be used separately. In this paper, the similarity of users was calculated according to their borrowing, so the similarity between readers and books was calculated with cosine similarity.
Variance analysis of the project-based similarity and user-based similarity: the above mentioned three similarity formulas are based on the scenario of project similarity, but there is actually a basic difference between project-based similarity and user-based similarity. To be specific, the calculation formula of project-based similarity solves similarity according to the similarity of column vector in the rating matrix, while the calculation formula of user-based similarity solves similarity according to row vector in the rating matrix. Therefore, the above three formula can be used separately. In this paper, the similarity of users was calculated according to their borrowing, so the similarity between readers and books was calculated with cosine similarity.
正在翻译中..