Recommender systems are programs that try to predict which objects ( movies , music , books , news , websites ) will be of interest to the user, having certain information about his profile .
Two main strategies for creating recommendation systems are content-based filtering and collaborative filtering [1] [2] . When filtering based on content, user and object profiles are created, user profiles can include demographic information or answers to a specific set of questions, object profiles can include genre names, actor names, artist names, and other attribute information depending on the type of object. For example, in the Music Genome Project, a music analyst evaluates each track by hundreds of different musical characteristics that can be used to identify a user's musical preferences. Collaborative filtering uses information about past user behavior - for example, information about purchases or ratings. In this case, it does not matter with what types of objects the work is carried out, but at the same time implicit characteristics can be taken into account, which would be difficult to take into account when creating the profile. The main problem of this type of recommender systems is the “cold start”: lack of data on users or objects that have recently appeared in the system.
In the process, recommender systems collect user data using a combination of explicit and implicit methods. Examples of explicit data collection:
- request from the user to evaluate the object on a differentiated scale;
- requesting the user to rank a group of objects from best to worst;
- Presenting two objects to the user with a question about which one is better;
- a proposal to create a list of objects that the user likes.
Examples of implicit data collection:
- monitoring what the user is viewing in online stores or other types of databases ;
- keeping records of online user behavior;
- Tracking the contents of a user's computer.
Recommender systems compare the same type of data from different people and calculate a list of recommendations for a particular user. Some examples of their commercial and non-commercial use are given in the article on collaborative filtering . The interest graph is used to calculate recommendations [3] . Recommender systems are a convenient alternative to search algorithms, as they allow you to detect objects that cannot be found last. Curiously, recommender systems often use search engines to index unusual data.
Notes
- ↑ Y. Koren, R. Bell, C. Volinsky. Matrix Factorization Techniques for Recommender Systems // Computer . - IEEE. - T. 42 , No. 8 . - S. 30-37 .
- ↑ Recommender systems based on collaborative filtering, 2002 , p. 187.
- ↑ Recommendations based on a graph of interests .
Literature
- Melville P., Mooney R., Nagarajan R. Content-Boosted Collaborative Filtering for Improved Recommendations ( University ) // University of Texas, USA: Materials conf. / AAAI-02, Austin, TX, USA, 2002 .-- 2002 .-- P. 187-192 .
- Zhernakova O. Systems of recommendations and search for video content // Telemedia, 2012.
- Nadim Hossain. Why the Interest Graph Is a Marketer's Best Friend . Mashable. Date of treatment December 7, 2013.