Contemporary Recommendation Systems on Big Data and Their Applications: A Survey
Ziyuan Xia, Anchen Sun, Jingyi Jingyi, Yuanzhe Peng, Rui Ma, Minghui Cheng
Abstract
This survey paper provides a comprehensive analysis of the evolution and current landscape of recommendation systems, extensively used across various web applications. It categorizes recommendation techniques into four main types: content-based, collaborative filtering, knowledge-based, and hybrid approaches, tailored for specific user contexts. The review spans historical developments to cutting-edge innovations, with a focus on big data analytics applications, state-of-the-art recommendation models, and evaluation using prominent datasets like MovieLens, Amazon Reviews, Netflix Prize, Last.fm, and Yelp. The paper addresses significant challenges such as data sparsity, scalability, and the need for diverse recommendations, highlighting these as key directions for future research. It also explores practical applications and the integration challenges of recommendation systems in everyday life, underscoring the potential of big data-driven advancements to significantly enhance real-world experiences.