A Recommendation System Combining LDA and Collaborative Filtering Method for Scenic Spot
Shengli Xie, Yifan Feng · 2015
Researchers have long sought to find an effective and straightforward method to bridge the gap between us and big data. Especially during the big data era, how to find the needed information with rapid speed and exact result has become the central concerns of the internet users. This paper focuses on exploring the valuable data in UGC (User Generated Content), and recommending useful information to specified users. To achieve this goal, we model the social network, and then the LDA (Linear Discriminant Analysis), PCA (Principal Component Analysis) and KNN (K-Nearest Neighbour) algorithms are adopted to calculate the recommendation items. Our algorithm avoids the disadvantages of the common collaborative filtering algorithm that only behaviors is considered but without considering the behaviour results, thus our method effectively improves the accuracy of the recommendation system. Experimental results show that our algorithm improves the accuracy comparing with the CF algorithms.