In an effort to improve user experience, M4M, Malayala Manorama's match-making portal, has started using intensive data sciences.
Like any matrimony website, M4M originally provided several search and filter functionalities for users to scan for profiles and locate choicest ones. A user would have to sometimes browse through several pages to get to the optimum one. For attaining a marked improvement, M4M now uses all back data available to create an intuitive and intelligent recommendation system. This system uses machine learning to pick information from the activities of all users to generate interesting profile recommendations. It can identify preferences that are not explicitly specified. It learns and adjusts its algorithm continuously with every change in user browsing pattern. The system also monitors user filters and blocks actively to detect changes and make sure the results are most interesting.
On the technical side, this system uses advanced machine learning technologies. This is a hybrid collaborative filtering and content-based system.
The system saves time for the users and makes their profile search highly effective. There has been a good increase in communication effectiveness (views to contacts ratio) and user engagement. Users have reported very positive feedback about this feature as well.