Information Technology Reference
In-Depth Information
information in the user profile, the more likely it is to find related news for a user.
The news recommendation strategy is based on the recentness of the news as well as
the correlation of computed interests and their occurrence in the news. Based on the
defined architecture and introduced trackable user interactions, we demonstrate the
SERUM system in the next section.
7.2.1 SERUM Use Case—User Behavior Collection
In order to explain the interaction of the semantic tracking component with the news
recommendation system, we walk through the first and fourth trackable user inter-
action outlined in the previous section, and detail the resulting user model and the
recommendations. As mentioned in Sect. 7.2 , SERUM is a personalized news rec-
ommender where the user profile is created by tracking and analyzing user behavior.
Initially, after the first login, the user profile and the personalized news stream is
empty as depicted in Fig. 7.3 . The picture shows the empty user profile on the left
and the empty personalized news stream. To create the user profile, the user has to
interact with SERUM, to read news or to search for artists.
When the user starts reading, their first interaction is with a list of news articles
where they can choose what to read. The SERUM news list shows the article, an
Fig. 7.2
Visualization of the SERUM user tracking use cases
Search WWH ::




Custom Search