研究目的
To propose a framework of the CF recommender system based on various user data including user ratings and user behaviors, discuss key features of these two kinds of data, classify several typical CF algorithms as memory-based approaches and model-based approaches, and compare them through two case studies.
研究成果
The paper concludes that CF recommender systems can greatly help mobile Internet users find proper items without excessive time and energy consumption in the era of Big Data. It suggests further studies in handling massive data and developing distributed computing algorithms.
研究不足
The paper does not explicitly mention limitations, but potential areas for optimization could include handling massive data in time and developing algorithms for distributed computing systems.
