Personalized POI Travel Recommendation from Social Media

Travel based recommendation and journey planning are challenging tasks because of various interest preferences and trip restrictions such as limitation of time, source and destination points for each tourist. Large amount of data can be collected from the Internet and travel guides, but these resources normally recommend individual Point of Interest (POI) that is considered to be familiar, but they do not provide sufficient information to the interest preference of the users or hold to their trip constraints. In addition, the huge volume of information makes it a challenge for every tourist to pay attention to a potential set of POIs to make a visit in any unknown city. After the tourist discovers an acceptable set of POIs to go to, it’ll take abundant time and energy for him/her to make a brief outline of the suitable duration of the visit at every POI and the order in which to visit the POIs. To sort out these problems, an author topic collaborative filtering (ATCF) algorithm is suggested for personalized tours. This method suggests that the POIs are optimized to the users’ interest preferences and POI popularity. Hence, this method is elaborately explained here for tour recommendation problem based on similar user and similar city prediction, which considers user tags. It extends our method to provide personalized suggestions based on user geo co-ordinates points. In the first instance, the multiple users’ location histories are modeled using tree-based hierarchical graph (TBHG). Based on TBHG, HITS approach is developed in order to gather the interest level of a selected place and a user’s travel expertise (knowledge). Finally, HITS-based collaborative filtering technique is used to obtain GPS based personalized recommendation system. And for image based search similar images with the tag information are retrieved for the query image users have. Keywords— City prediction, GPS trajectories, Place of interest Travel Recommendation, User preferences.