ITM 760 Lecture Notes - Lecture 9: Megadeth, Recommender System, Last.Fm
Document Summary
Recommender system suggests megadeth from data collected about customer x. Examples include amazon, pandora, stumbleupon, netflix, del. icio. us, movie lens, last. fm, google news, youtube, xbox live, instagram (thru posts, or people). Shelf space is a scarce commodity for traditional retailers and: tv networks, movie theaters, . Web enables near-zero-cost dissemination of information about products. More choices necessitate better filters - recommendation engines (more of a need, then a nice" option). Recommend from the long tail based on popularities which is based on users, to head. Types of recommendations: editorial and hand curated. Example: globe and mail gives recommendations based on an editorial board. Example: news platforms recommend: simple aggregates, tailored to individual users. Amazon and netflix recommend based on purchases and based on what you view. Utility function (u): x (x) s = r: each customer can potentially rate each product, e. g. , 0 to 5 stars, real number in [0,1]