Applying Deep Learning to Recommender Systems - Making Smart Systems Smarter

Booz Allen Hamilton is applying Deep Learning to advance the state of the art for Recommendation Engines by fusing personalized collaborative recommenders with semantic understanding and relationships derived by deep learning algorithms. All sites strive to provide recommendation and personalization, however curating all the content for content based filtering is extremely laborious while collaborative filtering for a new site suffers from the "cold start" problem, as there is no history of user interaction to leverage. Advanced Deep Learning algorithms for Natural Language Processing allow the automated processing of the content of a site, producing semantic understanding, relationships and simple recommendations, paving the way for rich recommendations and personalization, derived from both the interaction and the actual content of the items.

Sumeet Vij is a highly accomplished Chief Technologist, IT Executive and Data Scientist with a proven track record of 16+ years in leading large scale complex IT projects form concept to production in the federal and commercial sector. He specializes in leading high performance teams in the arena of Data Science, Big Data, Machine Learning, Data Analytics, SOA, Semantic Web, BPM, Open Source projects and Enterprise Architecture. Sumeet is currently part of the Strategic Innovation Group of Booz Allen, leading multiple client engagements and advanced research projects focused on providing greater value and deeper insights from the large scale analysis of data. He is leading internal investments in the cutting edge area of Deep Learning in the field of Natural Language Processing and Computer Vision