Towards Development of a Hypertext Induced Topic Search Based Point of Interest Recommender System for Location Based Social Networks
John Vaseekaran S., N Srinivasan · Journal of Machine and Computing · 2025
Location-based social networks (LBSN) have a significant issue in the suggestion of points of interest (POIs) due to the sparsity of data, implicit input from users, and individual preferences. In most of the LBSN systems, there is no simple rating method for POIs, which is a major drawback for many users. Due to a lack of acceptable connections, such algorithms tend to provide a list of POIs that the user cannot consistently visit. There are many applications for the link data analysis, and the Hyperlink-Induced Topic Search (HITS) algorithm in particular, such as highest ranked search engine results predicated on the hyperlink configuration of the World Wide Web and analysing privacy in social networks in order to compute node weight and understand the elements of each object (endpoint) in the network. By using the HITS algorithm, we can promote POIs to LBSN users while simultaneously considering the influence of social ties. Our suggested model is tested on the Foursquare dataset and compared to the most recent POI recommendation algorithm. When we tested it against two prominent algorithms using real-world datasets, we discovered that our suggested approach performed better in terms of both variety and accuracy.