Psychometric Recommendation Engine (PRE)
Engine for Media Content
Media contents of all types are often purchased directly via the Internet. Films, songs, TV shows and more are downloaded or streamed to various end devices. Many providers of media contents give the user the option to receive (purchasing) recommendations tailored to their interests. Some examples of providers that use recommender systems are Amazon (films, TV series, music, books), Netflix (films, TV shows), Lastfm or Spotify (music). To generate appropriate recommendations, information about the user from various sources is aggregated. This information includes demographic information, specific genre preferences, or media titles that the user has purchased via the platform in the past. Often users are also asked to link their social media profiles (Facebook, Twitter, etc.) with the provider’s platform. The more information that is gathered about the user, the more precise recommendations can be given.
In a research project co-funded by the Commission for Technology and Innovation (CTI), the Media Psychology section team at the ZHAW, together with the Scuola universitaria professionale della Svizzera italiana (SUPSI) and the Swiss Distance Learning University of Applied Sciences (FFHS), developed a recommendation engine – commissioned by the Ticino company dixero – that can be implemented for various media contents (such as film, music, books, etc.). The special thing about it is that in addition to the usual indicators, such as demographics, interests or consumer preferences, psychometric variables are included to improve the quality of the recommendations. It was the task of the ZHAW Media Psychology section team to find out what psychological factors are associated with the choice of media contents and how they can be integrated into the recommendation engine.