Published Paper Details:

Digital information has grown exponentially with lot of choices for services and products. Hence, filtering, prioritizing and efficient delivering of relevant information to tackle the problem of information overload is needed. This problem is solved in recommendation systems by searching through huge volumes of dynamically generated information to provide users with personalized contents and services. Historical data of users� preference and their purchases to predict items that might interest the users are used by recommendation systems. Recommendation engines mainly use the classifier algorithms for best prediction. So, it is noted that to develop accurate classifier algorithm to enhance the performance of recommendation engines. Random forest and decision tree popular classifier algorithms of choice that shares merits of high accuracy, high classifying speed, strong learning ability and simple construction. In this paper, we analysis a system which uses Random Forest and Decision Tree at multilevel strategies to predict the recommendations based on skills while targeting users� mindset and current trends. Compared to decision tree that provides 84.2% accuracy, random forest is better with 90.1% accuracy provided with feasible performance.