PODCAST RECOMMENDATION ENGINE USING HYBRID RECOMMENDATION SYSTEM

Abstract

With the rapid growth of digital audio content, podcasts have become a popular medium for entertainment, education, and news consumption. However, users often face challenges in discovering podcasts that match their preferences due to the vast volume and diversity of available content. This study proposes a Podcast Recommendation Engine using a hybrid recommendation system that combines collaborative filtering and content-based filtering to provide personalized podcast suggestions. The collaborative filtering component leverages user behavior, such as listening history, ratings, and subscriptions, to identify similar users and recommend podcasts they enjoy. The content-based filtering component analyzes podcast metadata, including categories, descriptions, and keywords, to match user interests with relevant content. By integrating these two approaches, the hybrid system overcomes limitations such as cold-start problems and sparse user data, improving recommendation accuracy and user satisfaction. Experimental evaluation demonstrates that the proposed engine effectively personalizes podcast recommendations, enhancing user engagement and discovery of relevant content.