When a subclass of the information filtering system can predict the preferences or ratings that a user would give to an item or like, it is called a recommendation engine system. They are mostly used in the case of commercial applications.
They are used in a wide variety of areas, but most commonly recognized as playlist creators in various music services like YouTube, Spotify and even as Video or movie preferences in Netflix, Amazon Prime, etc. they are also used for product recommendations in online sites like Amazon and Flipkart and also contains recommendations in Instagram, Facebook, Twitter.
These systems operate either using a single input, like music, or multiple inputs within and across platforms like news, books, and search queries. Online dating and restaurant sites have specific popular recommendation engines. To further explore research articles and experts, collaborators and financial services recommendation engines have also been developed.
The recommendation engines mainly make use of both collaborative filtering as well as content-based filtering (personality-based approach), also other systems such as knowledge-based systems.
The Collaborative filtering approach builds a model from a user’s past behavior, for example, like, items previously purchased or selected as well as similar decisions made by other users. This model is then made use to predict items that the user may show interest in. whereas, the Content-based filtering approach utilizes a series of discrete, pre-tagged characteristics of an item in order to recommend additional items with similar characteristics. However, the current recommendation engines typically combine one or more approaches into a hybrid system.
In 1990, at Columbia University, the Recommendation engines/systems were first mentioned by Jussi Karlgren as a “digital bookshelf”. From then it was implemented at scale and worked through in technical reports and publications.
A distinction is mainly made between explicit and implicit forms of data collection while building a model from a user’s behavior,
Examples of explicit data collection include the following:
- Asking a user to rate an item on a sliding scale.
- Asking a user to search.
- Asking a user to rank a collection of items from favorite to least favorite.
- Presenting two items to a user and asking him/her to choose the better one of them.
- Asking a user to create a list of items that he/she likes
Examples of implicit data collection include the following: Observing the items that a user views in an online store, Analyzing item/user viewing times, Keeping a record of the items that a user purchases online, Obtaining a list of items that a user has listened to or watched on his/her computer or Analyzing the user’s social network and discovering similar likes and dislikes.
Whereas, content-based filtering mainly makes use of the data like a model of the user’s preference and the user’s search history.
In this article, we have mainly discussed the main two types of Recommendation engines. There are also other varieties like Multi-criteria recommendation engines, Risk-aware recommendation engines, Hybrid recommendation engines, Mobile recommendation engines, etc.