Analysis of Final Ratings in Tourism Using Various Machine Learning Techniques

Abstract

Abstract: User-generated ratings are a tremendous asset to product descriptions and significantly impact decision-making in Context-Aware Recommender Systems. Researchers exploit this information to predict user preferences, model the item's attributes, and offer intelligible recommendations. However, not all contextual ratings are significant because they may be posted by various users for various reasons and based on different routines. Further, as users care about different attributes of multiple contexts, not all user ratings equally reflect the users' opinion of the overall rating, a primary concern in recommender systems. This article predicts the overall rating using user-user and item collaborative filtering with significant contexts and the outcome tested on various machine and deep learning models using the contextual segments.