Diligently working behind the scenes of all OTT Web sites is something called a recommendation engine. Users normally take no notice of it but it is constantly interacting with them. It yearns to learn our taste in music, understand our preferences in movies, and predict what we might want to buy. Even the simplest task of finding an interesting Web site to enjoy is the result of a recommendation engine. Who did you think recommended that obscure but now favorite movie you just watched on Netflix or LoveFilm?
Subtle and silent, the recommendation engine is making long strides to enter every part of our lives. Don’t be alarmed! Recommendation engines are desperately needed especially now because there are so many content-rich OTT sites.
Searching fruitlessly is difficult and time-consuming when a user doesn’t know what to search for. Browsing is more effective, but who has the patience? Browsing is child’s play for a recommendation engine. These engines are designed to read our mind, know our soul, and unlock the desires of our heart — at least for online entertainment. One such engine that is slowly but steadily making itself useful is the UK-based The Filter.
The Filter is a recommendation engine with some big name customers –Nokia, Dailymotion, NBC, Warner Bros. and Walmart’s Vudu, to name a few.
The Filter recommends music, movies, Web-based videos and Web pages to users for free. This system knows not only what the user watched or listened to but also whether they stopped enjoying a product halfway through. This allows The Filter to not only recommend a popular song or video but also recommend something off the radar. The Filter is a hybrid recommender. At the core it uses item-based collaborative filtering, but overlays an individual user’s taste graph to provide highly individual recommendations. It is also sensitive to the context (time of day, location, day of week, device type) and uses this to filter recommendations to ensure they are even more relevant.
The SaaS (Software as a Service) platform that The Filter uses complies with all types of digital content services such as IPTV, mobile, music, movie and Web videos. Therefore, recommendations are possible across all types of media. All of this is achievable through the use of flexible application program interfaces (APIs).
Despite everything that The Filter has achieved in relevant and personalized recommendations, there are still barriers. To become a popular and well-used application on lots of Web sites, The Filter must discover a way to scale down and reach out to small and medium-sized businesses as well as be sold at an attractive price.
The Filter is just one of several recommendation engines we will continue to cover, each with its own strengths and weaknesses.
This article has been edited from its original version.