Personalized Information Access Using Semantic Knowledge - Smart Information Systems: Computational Intelligence for Real-Life Applications - page 179

which Carl can easily answer. One fact all of them notice is the information quality.
Carl had just started using SERUM, but the app already provides him with relevant
information regarding his information need. Other apps need far more interaction
for that.
A few days later, all friends have downloaded the app, SERUM causes the next
round of discussions on the schoolyard. By using SERUM, almost everyone now
gets news about their favorite band. But SERUM shows more than pure news. It also
gives information about related content, such as bands similar to the one currently in
the news. As a result, everybody finds some new bands, which are in their personal
opinion better than the others. This leads to endless discussions before, during, and
after school among Carl's friends. Carl stays out of all these discussions; he is still a
fan of the band starting all that fuzz and he cannot wait to go to their concert tomorrow.
7.1 Introduction
The flood of available information and products offered by Web applications like
online retailers and news portals overwhelms today's users. To handle this informa-
tion overload, applications typically offer some kind of personalization techniques,
in most cases, in the form of personalized filtering or personalized recommendations
[ 3 , 36 ]. However, personalized recommendations that adapt to the users' individual
taste are a major challenge [ 1 ]. On the one hand, personalized recommendations
improve user satisfaction and can motivate users to return. Bad recommendations,
on the other hand, may cause users to turn their back on those applications. A com-
mon recommendation approach is Collaborative Filtering (CF). CF utilizes histori-
cal user information, like ratings or interactions, to compute recommendations [ 37 ].