The Research On The Recommendation Technology Of E-commerce Based On LBS

With the rapid development of mobile Internet and mobile commerce, as one ofthe core technologies about e-commerce, the personalization recommendationtechnology is also becoming the hotspots in the field of mobile commerce. However,there is less study on the personalization recommendation technology of mobilecommerce, and the existing research results are far from the solution to all problems.So, based on the traditional recommendation technology and the characteristics ofmobile commerce (location-relevance), this paper studied recommendationtechnology about mobile commerce. Main work and conclusions：First of all, this paper presents an L-U-I model based on LBS. In this model,location as a new dimension is introduced into user interest model, and the userinterest is represented as a point in space. Secondly, based on the model,computational method about the user interest and the local similarity are improved, anew personalization recommendation algorithm is put forward. The main idea is thatthe algorithm discover the closest locations by calculation on the local similarity，andthen the top ordination of the item by calculation on user interest will recommend tothe user. Finally, this paper has designed a series of experiments to test and analyzethe algorithm. The experiments recommend suitable restaurants for users in mobileapplication environment.The experimental results verify that the personalized recommendation algorithmbased on the LBS can meet the need of mobile commerce application. It is sensitive tochange about the location and the user interest, and also can handle the big changesabout mobile environment. In addition, the accuracy and response time of therecommendation algorithm results the effectiveness of the algorithm.