Adaptive Tourist Attraction Recommender: A User-Centric Approach Utilizing Collaborative Filtering and Cosine Similarity

Abstract

This paper presents the development of a user-based tourist attraction recommender system designed as an interactive online application. The primary objective of the system is to generate a personalized list of recommended tourist attractions tailored to individual user preferences. The system leverages well-established techniques from classical recommender systems, particularly collaborative interring, to adapt effectively to the tourism domain.

The recommendation process is structured into three core phases: (1) representation of user information, (2) identification of similar (neighbour) users, and (3) generation of attraction recommendations. To accurately measure user similarity, the Cosine similarity method is employed during the neighbour generation phase. Subsequently, recommendations are derived based on the visitation history of users with similar preferences. To illustrate the functioning of the system and the underlying recommendation algorithm, a detailed case study is provided. This demonstrates the practical implementation and effectiveness of the proposed model in delivering location based, user-personalized tourist suggestions.