Research on entropy-based collaborative filtering algorithm and personalized recommendation in e-commerce

Abstract

Based on the introduction to the user-based and item-based collaborative filtering algorithms, the problems related to the two algorithms are analyzed, and a new entropybased recommendation algorithm is proposed. Aiming at the drawbacks of traditional similarity measurement methods, we put forward an improved similarity measurement method. The entropy-based collaborative filtering algorithm contributes to solving the cold-start problem and discovering users' hidden interests. Using the data selected from Movielens and Book-Crossing datasets and MAE accuracy metric, three different collaborative filtering recommendation algorithms are compared through experiments. The experimental scheme and results are discussed in detail. The results show that the entropy-based algorithm provides better recommendation quality than user-based algorithm and achieves recommendation accuracy comparable to the item-based algorithm. At last, a solution to B2B e-commerce recommendation applications based on Web services technology is proposed, which adopts entropy-based collaborative filtering recommendation algorithm.

Popularity and user diversity of online objects

“…In addition, we have not investigated the constitution of the types of users who select the objects with specific degree, which is a still an open question. Furthermore, it is important to verdict whether the information entropy could be considered in the recommender algorithms [30][31][32][33].…”

A review of movie recommendation system: Limitations, Survey and Challenges

“…It is used to recommend any item based on related item's rating. By analyzing the rating , the users whose rating are similar for different item, only those items are available for recommendation [36]. It has been widely used by all Web giants including Netflix, YouTube etc.…”

A Collaborative Filtering Algorithm Combined with User Habits for Rating

“…As is known that in a thermodynamic system, entropy is a measure of d isorder or chaos in the universe. Shannon defined entropy of information as the uncertainty of source by analogy [8][9][10] . Assuming that the source has n kinds of possible values, the corresponding probability for each value is 12 , , , n p p p , entropy of in formation is defined as: As Table III. shows that, a certain user with higher Shannon entropy has relatively decentralized rat ing distribution.…”