Shopping Mall Recommended System Research And Application Based On LBS

At present, the shopping mall’s shopping-mode is still one of the main business sales model, study market environment of mobile service become the core of the market competition ability. This paper proposes a location based services recommendation system for the shopping mall, in order to enhance the personalized customer service ability, provide accurate service.Firstly, this article detailed analysis the demand of the recommendation system on the shopping mall, this paper expounds the mall recommendation system in mobile environment with user mobility, short-term interest sensitive, recommend precision demands higher. The article puts forward the recommendation system design based on the store location service. The location will be introduced as a new dimension to the recommendation model, this paper generate a "user-goods-location" recommend 3d model, and through the semantic position to improve the traditional recommendation model in mobile environment.This article focuses on the recommendation algorithm based on the location of services research. The algorithm uses spatial density clustering algorithm analysis the user’s location tracks, get user’s interested in a collection of location area of the shopping mall, combined with commodity classification and location information, converted into the user’s geographic location to mall semantic position. Then, divide different users through the semantic similarity algorithm. However, the recommended method is not optimized for large computation and accuracy is not high, we use the recommended method to location-based services to achieve the goal, to improve accuracy and reduce Small-scale computing. This method is mainly used to calculate the similarity between users semantics position vector, amendment user rating matrix through space semantic positional relationship mall. Then the amendment scoring matrix using the evaluation preferences, and determines user ratings trend for commodity and commodity trends are scored to improve ratings sparse matrix problems. Finally, using filtering collaborative algorithm calculation rating similarity and sorting, then produce the K nearest neighbor users of the target user. By the nearest users, we can generate products to be recommended by the nearest neighbors, and pushed to the user’s mobile terminal via a network and at the same time we collect user’s feedback to improve the algorithm.Application results show that the constructed LBS mall recommendation system is good, it can satisfy the needs of shopping malls recommend moving environment, greatly improve the shortcomings of traditional recommendation algorithm in the mall environment, through the system recommend ranking score and generating the shopping-set time show that the recommended method has high accuracy and also recommended faster. While addressing the impact of user ratings sparse matrix algorithms and other issues, it has a certain application value.