In the era of World Wide Web, where the number of choices is irresistible, there is need to prioritize, filter, efficiently and effectively deliver significant information to solve the problem of information overload, which has posed a potential crisis for many Internet users. Recommendation systems answer this problem by penetrating through voluminous dynamically produced data to offer users with personalized information and services. This paper puts light on the various features and existing power in different prediction methods of recommender systems for assistance in research and practice in the area for developing powerful recommendation systems.