ARS, articles recommendation system
ARS is an innovative system of textual knowledge management working in semantic domains. It uses Neodata proprietary algorithms to select articles by semantic similarity and to recommend them to online readers in real-time sessions.
ARS performs three levels of recommendations:
- Semantic Recommendation: select all articles related to a specific one according to their content affinity;
- Semantic Recommendation combined with Collaborative Filtering: collect preferences from a large number of users to build up associative rules. This technique considers both content similarity and associative rules to recommend the most relevant articles.
- Semantic and Collaborative Filtering combined with Profiling Recommendation: make the most of content affinity, users’ preferences, and user profile. ARS uses Neodata system of behavioural targeting to build up a profile for each user.
When reading an article, ARS provides the online reader with a list of recommended articles for raising the user interest and encouraging him/her to continue to read online.
Repubblica.it uses ARS to manage the following three boxes on is portal:
- "Sullo stesso argomento" is the ‘on the same topic’ box: ARS produces a list of articles according to their semantic similarity with the article that the user is reading online
(eg.: http://www.repubblica.it/politica/2010/05/28/news/fini-dittatura-4398160/); - "Leggi anche"is the ‘read also’ box: ARS produces a list of articles combining their semantic similarity with the article that the user is reading and articles’ preferences collected from a large number of users (collective intelligence). If many people interested in article A have read also article B, as soon as a user exhibits some interest for A the recommendation engine automatically suggests B
(eg.: http://www.repubblica.it/2009/03/sezioni/politica/dl-sicurezza-3/cavaliere-irritato/cavaliere-irritato.html). - "I più letti" is the ‘most read articles’ box (Top News): ARS produces a list with the most read articles in the last week
(eg.: http://www.repubblica.it/politica/2010/05/28/news/fini-dittatura-4398160/).
