A Hybrid Approach for Recommender Systems in a Proximity Based Social Network

Being one of the latest trends in technology, big data is proving to be fundamental in various fields and domains. Analyzing the large volume of data leads to fruitful information and depicts new methods of achieving growth and innovation in this competitive world. Similarly, analyzing large data sets from social media can enhance recommendations provided by recommender systems in a proximity based social network. This research work presents a hybrid approach for performing recommendations in a proximity based social network by using three recommendation techniques namely Content-based filtering, Collaborative filtering and Link Analysis. Additionally, big data from social media is analyzed to enhance the recommendations. The Hadoop ecosystem is used to help for processing large datasets. A prototype has been implemented and evaluated.