Personalized Novel Recommendation System Based on Filtering and Sentiment Analysis

In recent years, the rapid advancement of Internet technology has led to the emergence of online novel reading as a prevalent leisure pursuit. Reading literature online can result in a substantial amount of data, potentially causing data overload for individuals. To tackle this challenge, Recommender Systems (RS) have been introduced, which play an important role in selecting selective information. Users product reviews influence whether users decide to read the novel or not. User decisions and beliefs are frequently influenced by the responses exchanged amongst other customers on the internet, blogs, and social media platforms. Coordination filtering and content-based recommendation are the two fundamental filtering techniques that make up RS, but there are some limitations such as cold start and users’ historical behaviour and preferences. Therefore, this research suggests a hybrid recommendation system that incorporates sentiment analysis together with content-based recommendation and coordinated filtering. Tests were carried out on publicly available datasets and shown superior accuracy in comparison to cutting-edge techniques.