Recommender Systems: Harnessing user-generated content to drive business growth
When shopping online, posting photos or messages on social media websites or posting product reviews, users generate a tremendous amount of data that reflect their interests, opinions and habits. Businesses of all sizes have already such data generated by their internal systems and blogs, or could gather this data from public websites. This data is used to boost product marketing and grow sales using recommendation system techniques.
This talk will show how recommendation systems are built in industry to address different business requirements. Recommendation approach selection is discussed with regard to the type (e.g. transactional, textual, ratings,…) and scale of the available data. A worked example of the implementation and validation of a recommendation system for an online retailer is presented in Python. Experiment design to test recommendation systems in production will be also discussed.
Principal Data Scientist, Insights Centre
Houssem Jerbi is a data science and analytics consultant. He is currently working as consultant at Facebook, Dublin. Previously, he was Research Scientist and Technical lead at Insight Centre for Data Analytics. During his work in both R&D and engineering, he has led high-impact projects in data science and business intelligence across different sectors, including Retail, Digital & Social Media, Advertising, Hospitality and Real estates. Houssem holds Ph.D. and M.Sc. degrees in Computer Science from University of Toulouse, France.