Personalized Recommendation of Movies Using a Combined approach of locality sensitive hashing, K-Nearest neighbour and collaborative filtering.
Chathuri Madhushika · International Journal of Research Publications · 2018
With the highly expanding usage of social media, a huge amount of data is being collected day by day. Proliferation of movies and their related views on social media has posed significant challenges in discovering the most suitable movies. Personalized recommenders that have been implemented to recommend movies use only a source of data and largely overlooked integration of several big data sources including tweets, you tube comments, posts on movies and comments. Moreover, most of the current approaches focus on one single aspect i.e. either content-based, collaborative filtering. Overlooking the integration of multiple aspects with many data provenances caused low effectiveness in current approaches. Hence, to overcome these deficiencies a novel, hybrid approach, KLC model which integrated user-based collaborative filtering, locality sensitive hashing and network-based approach is proposed. The New KLC model has been tested with 10000000 data items it outperforms benchmark methods