Recommendation systems are a type of information filtering systems that recommend products available in e-shops, entertainment items (books music, videos, Video on Demand, books, news, images, events etc.) or people (e.g. on dating sites) that are likely to be of interest to the user.
Gravity’s recommendation system automatically learns and analyses the browsing / shopping behavior of the user on a given website or platform. Then based on this information and based on several other similar profiles the recommendation system shows items that are fit the taste of the given user.
Gravity’s recommendation system uses collaborative filtering (described above) and matrix factorization (pdf) techniques – some of them patent pending – in order to learn the taste of the user, thereby constantly enhancing the quality of recommendations.
When building the user's profile a distinction is made between explicit and implicit forms of data collection. Explicit user data includes ranking, liking, or favoring a item while implicit data refers to the viewing time, number of views, and actual purchases of an item.
The recommendation system then compares the collected data to similar and not similar data collected from others and calculates a list of recommended items for the user.
Recommendation systems are a useful alternative to search algorithms since they help users discover new items they might not have found by themselves.
Implementing recommendation systems into e-commerce platforms significantly enhances customer satisfaction thus increases average basket value and customer lifetime values.
The effectiveness of Gravity’s recommendation system was proven by the Netflix Prize contest, where thousands of companies, scientists and university research groups were competing to enhance the quality of algorithms. Gravity’s team, as a part of the Ensemble was ranked second at the end of the 3 year competition by enhancing the recommendation quality by 10.1%.