Gamification of Workplace Activities - Smart Information Systems: Computational Intelligence for Real-Life Applications - page 255

behavior in a gamified environment [ 22 ]. Recent studies also indicate that the effect
of game design elements can change over time, which can end up with lower effects
in the long term [ 8 , 10 , 12 ]. In the worst case, positive effects might only be caused by
the novelty effect. In the next section, we outline how player types could be identified
automatically, hence reducing the risk of triggering negative gamification effects.
9.4 Towards the Automated Identification of Play-Personas
Existing gamification definitions pursuing the increase of user experience [ 6 ] and
overall value [ 15 ] indicate that the application of gamification is goal oriented.
Therefore, we usually look at gamification as the necessity to maximize an over-
all goal . However, the rich variety and individuality of the users results in different
behaviors, preferences, and motivating factors (e.g., [ 11 , 17 , 35 - 37 ]). In the worst
case, negative effects can occur when applying gamification, as observed by Hamari
et al. [ 13 ] and Mosca [ 25 ]. Hence, for successful gamification, several factors needs
to be considered, which makes the design process difficult and expensive. Therefore,
our extended look at gamification is the necessity to maximize an overall goal with
respect to the individuality of users .
In this section we propose a new approach for gamification based on the automatic
detection of play-personas. Dixon et al. [ 7 ] consider play-personas “as a useful tool
that can be used to put player type research into practice as part of the design process
of gamified systems.” In order to automatically determine different personas, we need
to reduce the effort to determine relevant player types for implementing gamification.
Why not skip the determination of player types and directly suggest game design
elements? Trying to achieve this with questionnaires and interviews can of course
increase the design effort. However, what if a formula or tool that helps to select
such game design elements based on experiences learned from user interaction data
over time can be used instead? Under the assumptions that (i) gamification consists
of various types of users that experience game design elements differently; and (ii)
gamification is deployed in order to achieve some goal in the broadest sense, we pose
the gamification design problem as that of assigning each user (at least) one game
design element that maximizes their expected contribution in order to achieve that
goal.
We suggest matrix factorization to create a genericmodel based on user interaction
data as a suitable methodology which could help for the selection of most fitting game
design elements. Parts of the treatment are based on [ 30 , 33 ]. The hypothesis is that
predictive models as intelligent tools for supporting users in decision-making may
also have potential to support the design process in gamification. We argue that this
not only reduces the design effort, but also provides a better selection of game design
elements since this kind of selection would not only be based on how users perceive
gamification [ 22 ] but also on their actual interaction with game design elements. We
are convinced that such data-centric tool can support the design process substantially.