Personalization and Recommendation Services - Smart Information Systems: Computational Intelligence for Real-Life Applications - page 122

Although a multitude of new movies are released every year, the frequency of
these releases is far lower than it is in other domains. In Chap. 6 , Kille et al. present
the use case of online news recommendation that differs signi

cantly from the
movie scenario. In the news domain publishers constantly provide new news
articles, resulting in vast amounts of items and a constantly changing dataset.
Besides, freshness is an important aspect in news recommendations. While users
may appreciate movie recommender systems suggesting movies from decades ago,
they will most likely not be interested in receiving yesterday s news as a recom-

mendation. Moreover, news publishers often have very limited knowledge about
their readers, i.e., recommendation algorithms have to deal with incomprehensible
as well as inconsistent user pro

'

les.
Focusing on a hybrid recommendation technique, Plumbaum and Lommatzsch
showcase in Chap. 7 how knowledge about individual users can be exploited to
provide recommendations. Focusing on the news domain, they outline a system that
provides news from the entertainment

preferences. In order
to capture these preferences, they introduce an ontology-based user behavior model
and present an evaluation that showcases the bene

eld that match users '

ts of using such approach.
Recommendation services have successfully entered online retailing businesses.
Online retailers rely on their ability to direct users to products they will enjoy. In
Chap. 8 , Plumbaum and Kille investigate speci

cities of recommending fashion to
men. Use cases mentioned before target consumable products. Users watch movies
and read news articles. In contrast, systems recommending fashion articles are
subject to further restrictions. Recommended products not only have to appeal to
users but also match to pre-existing items in their wardrobe. In their chapter, the
authors introduce strategies to deal with these additional requirements.
In the last chapter of this part, Chap. 9 , Meder et al. illustrate how gami

cation can
be applied to motivate users to provide manual recommendations. Gami cation

refers to the use of principles borrowed from computer gaming to increase user
engagement. They embed their scenario in an of ce workspace environment where

various IT systems exist that are designed to share knowledge between employees.
They argue that such systems are often not used by employees and suggest to
gamify these systems, hence providing the means to increase users

activities.
Furthermore, they propose a methodology to identify different types of employees,
referred to as player types, which allows them to adapt the use of gami

'

cation

elements based on the preferences of these player types.