Game Development Reference
In-Depth Information
With these results, we consider the approach to be delivering valid tags in general.
From the correct tags, the judges identified 38.5% to be specific. Here, the groups
also differed: the purpose aware group players provided 44% and unaware group
33% of specific tags. Considering this, we can conclude, that our approach is also
able to acquire the valuable specific information for personal archives.
Overall, we can conclude, that the PexAce-Personal gameplay was capable to
deliver valid metadata for the featured images, partially also specific for the needs
of personal image repositories. The results also shown that all players were able
to deliver a valuable output regardless of the awareness about the game's purpose,
however, we have also observed quantitatively and qualitatively larger contribution
from purpose aware players. Only in case of understandability the “unaware” group
performed better (which was according to our expectations). A typical case of correct
specific but “not-understandable” tag was a name of an event to which some images
were related. From this, we conclude, that making the players aware about the game's
purpose makes a significant positive impact on the quality and quantity of acquired
tags. We attribute this to the combination of this awareness with the motivation of
“working for oneself”. On the other hand, the purpose aware players reported that
they were also less enthusiastic about the gameplay and the focus on the providing
better annotations distracted them a bit from the gameplay fun.
The follow-up analysis shown that the specific tags contained (were of type):
53% person names,
21% event,
15% places,
11% other.
The “other” typed tags mostly have humorous connotations, as reported by judges.
This shown, that the players like to entertain with the annotations themselves. The
humor is probably also the reason for relatively outstanding difference between
tag specificity levels delivered by players of different groups for situational image
category: while for the purpose unaware players, this was a perfect occasion to
make jokes in the annotations, the purpose aware group was more disciplined. The
groups were much more coherent (in terms of specificity) in case of portraits and
group photos. The worst specificity results were achieved for images not containing
persons. This suggests that in future use, the game might utilize an automated method
for person presence detection in the images loaded to it, in order to increase its
usefulness, in terms of tag specificity.
Lastly we have qualitatively analyzed the session records, namely the player
verbal expressions about the game. As most important, two notions were recurring.
As a positive outcome, the players reported, that they enjoy playing with their own
images, especially longer unseen. On the other hand, players were skeptical about
the repeated use of the same images in multiple games player shortly after each other.
They were confused during the gameplay, unsure whether the image they have just
seen, they recall from the actual or past game. They have also reported, that they tend
to use exactly same annotations for them(which may cause a lack of tag diversity
Search WWH ::
Custom Search