A Spatial Distance Adaptive Next Interest Point Recommendation Method

Abstract

The invention discloses a space distance adaptive next point of interest recommendation method. This method combines the Markov chain and can adapt to the user's personal preference to recommend points of interest. A Markov chain is used to capture the temporal relationship of user check-in sequences, and a personalized transition matrix is ​​generated for each user to capture user preferences. The tensor decomposition model is introduced to solve the problem of data sparseness in the data set, so that each transfer matrix is ​​affected by similar users, similar points of interest, and potential preferences of users, and a more complete and high-quality user transfer matrix is ​​generated, which is captured by this transfer matrix. Perfect the transfer relationship between users' personal preferences and points of interest. Learn the user's personalized potential behavior pattern through the user's check-in sequence, so as to capture the effective user's potential personal preference. By fusing the spatial distance, the user's long-term potential personal partial information is chosen, and the user's personal preference is adaptively adapted according to the distance between the points of interest.

Description

technical field [0001] The invention belongs to the field of recommendation systems, and in particular relates to a method for recommending the next point of interest according to the spatial distance self-adaptive user interest and point of interest transfer relationship. Background technique [0002] In recent years, with the continuous development of network technology, people can obtain more and more information on the Internet, and it has become a huge challenge to accurately recommend information that users are interested in from massive data. In response to the problem of information overload, researchers have proposed a variety of recommendation systems for music, movies, advertisements, commodities and other fields. For example, in Netease Cloud Music, similar music will be recommended based on the music users often listen to, while on JD.com and Taobao, products will be recommended based on the products purchased by the user in the history and the products that the...

AI Technical Summary

Problems solved by technology

[0005] Aiming at the deficiencies of the prior art, the present invention proposes a space-distance self-adaptive next point-of-interest recommendation method, combining user check-in sequence, context information and spatial relationship between points of interest, using Markov chain and matrix decomposition method Solve the problem of sparse check-in data for check-in points of interest, and finally sort the user's prediction results by Bayesian sorting, and recommend the next point of interest to the user

Images