Identification Of User Preference by Sequential Pattern Mining and Recommendation of Products using Geo-tagged data

Abstract

Recommendation System provides the user with the interesting materials which are extracted from their preference. For simplifying the information retrieval and in order to provide the user with preferred result with more accuracy, recommendation system is being used. Recommendation based services are also used in social networks such as Facebook, Twitter, Instagram etc. Geo-tagged data plays a major role in case of recommendation systems as they will be providing recommendations with respect to the users locations. The semantic classification of the location is done using Support Vector Machine. By considering the location co-ordinates the nearest possible travel routes are identified by Google Maps and the shorter distance are computed using k-Nearest Neighbour. In this work, recommendation of products is given by means of considering the frequent buying pattern of the user using Prefix span algorithm, similar users ratings computed by Collaborative Filtering and the popular items available on the travel route. The proposed system has been implemented and evaluated.