Recommendation Systems for Mobile Devices

Recommendation Systems for Mobile Devices

Fabrizio Caruso, Giovanni Giuffrida, Diego Reforgiato, Calogero Zarba
DOI: 10.4018/978-1-4666-1981-4.ch014
(Individual Chapters)
No Current Special Offers


The authors describe three recommendation systems for online articles that are specifically tailored for mobile devices. In order to increase the number of articles read by the average user, an online newspaper could be personalized for each reader. Each user receives a personalized selection of the articles that take into account the limited bandwidth and screen, the user’s preferences, and possibly their geographical position. Two general criteria are followed: a collective intelligence criterion and a content similarity criterion. The suggested articles need to be both popular among the members of the online community, and similar to the articles already read by the user. The three systems address three similar problems. NeoPage is a tool for newspapers’ editors that suggests the position that each article should have on a web page. ARS is a tool for newspaper readers, which recommends the most similar articles to an article just read. MyNews is a tool for the readers, which produces a list of recommended articles by taking into account both the popularity of the article and the previously read articles by the user.
Chapter Preview


As mentioned in the previous section, two different criteria are followed in article recommendation systems: collective intelligence and content. The former depends on some explicit or implicit rating given by the users; the latter depends solely on the words contained in the article.

Several recommendation systems have been proposed where different approaches are used (content-based: NewsWeeder (Lang, 1995), Krakatoa Chronicles (Bharat and Kamba and Albers, 1998), PersoNews (Banos et al., 2006) ; collaborative filtering: Google News (Das et al., 2007).

As far as the three systems presented here are concerned we have that NeoPage is a tool for online newspapers that uses collective intelligence; ARS is a primarily content-based recommendation system; MyNews is a hybrid system.

The common goal of the three systems we present here is to increase the total number of articles' visualizations, which we call “impressions”.

Complete Chapter List

Search this Book: