Recommended system and method with facing social network for context awareness based on tensor decomposition

Abstract

The invention discloses a recommended system and a method with facing social network for context awareness based on the tensor decomposition, and relates to the field of the data mining and the information retrieval. Firstly, the method makes use of a social network massive data set to collect users and projects and contexts, to pay attention to the list information, to establish an original the user-the project-the context mark matrix, to calculate the users similarity, and to establish a user-user similarity matrix; Secondly, aim at the extreme sparsity of the original mark matrix, a sparse mark matrix is predicated and filled by using the tensor decomposition; Thirdly, aim at a problem that the user similarity matrix is sparse, a sparse user similarity matrix is predicated and filled by using the matrix decomposition; Finally, according to some similar interest tendencies of some similar users in the social network, a social normalization item is taken to optimizing the mark matrix. The method deals with the problem that a traditional predicated mark matrix does not consider that the context information and the relationship between users have an effect on marking. Also, the method deals with an obstruction which is caused by the sparsity of the mark matrix brings to the recommended system, thus the accuracy of the recommended system is improved. The method can be widely applied to the fields of the social network, the electronic commerce and the like.

Description

Technical field [0001] The invention belongs to the field of data mining and information retrieval, relates to personalized recommendation of a recommendation system, and is a context-aware recommendation method for social networks based on tensor decomposition. Background technique [0002] With the rapid development of the Internet, people are gradually entering the age of information overload. To solve the problem of information overload, people continue to adopt new measures, such as strengthening search engines, optimizing recommendation systems, etc. to solve the problems caused by massive information. [0003] In recent years, recommendation systems have been favored by IoT giants and e-commerce companies. In particular, the development of personalized recommendation technology has played an important role in improving user experience and improving service quality. The current main recommendation methods include content-based recommendation, collaborative filtering recommen...

AI Technical Summary

Problems solved by technology

The content-based recommendation system first extracts the content features of the recommended object, and then matches the product features with the user's interests and preferences, and recommends products with a high degree of matching to the user. Since there is no effective feature extraction method for multimedia resources, content-based The recommendation system will be limited by the feature extraction ability of the recommended object; for collaborative filtering recommendation, first find the nearest neighbor set similar to the target user's interest preference, and then predict the score of the target user's unrated item based on the score of the nearest neighbor set on the item , select the N items with the highest predicted scores as the recommendation results to feed back to the user. This recommendation algorithm can discover the user's undiscovered and potential interest preferences. However, the collaborative filtering recommendation algorithm still faces cold start, sparsity, accuracy, and expansion. issues such as sex; the recommendation based on association rules is based on the association rules between items, and the potential association between items is found through data mining to implement joint recommendation. However, when the amount of data is very large, the computational complexity of this recommendation algorithm It will increase accordingly; the hybrid recommendation system achieves the purpose of maximizing strengths and avoiding weaknesses by combining different recommendation strategies, so as to generate recommendations that are more in line with user needs. However, the hybrid recommendation system faces many difficulties in practical applications. It needs to solve different recommendations Technical Difficulties Making Organic Recommendations

[0004] To sum up, with the sharp increase in the number of users and items (items), traditional recommender systems face enormous challenges, including the cold start problem, the sparsity problem of the rating matrix, etc., especially when the original user-item rating matrix is ​​high The sparsity seriously affects the recommendation quality

At the same time, most of the traditional recommendation algorithms only consider the single factor of rating, without combining the knowledge of social network analysis theory, ignoring the user's social relationship, item association attributes and some contextual information (geographical location, emotional factors). To some extent, the accuracy, novelty and coverage of the recommendation are reduced

Images