Show simple item record

dc.rights.licenseCC-BY-NC-ND
dc.contributor.advisorStrauch, C.
dc.contributor.authorMohammad, F.
dc.date.accessioned2021-08-12T18:00:20Z
dc.date.available2021-08-12T18:00:20Z
dc.date.issued2021
dc.identifier.urihttps://studenttheses.uu.nl/handle/20.500.12932/40748
dc.description.abstractBackground and motivation. Regular eye-tracking studies rarely report data from more than a hundred participants. This is mostly due to the considerable effort involved in assessing data, given that usually only one participant can be tested at a time and an experimenter needs to be present and eye-tracking equipment is relatively expensive. In the present thesis, I present data obtained from an eye-tracker that was available to the public at the NEMO science museum Amsterdam. As part of a display at the museum, gaze data from more than a thousand participants was assessed. This is a unique opportunity for improving the breadth and depth of behavioral studies while taking into account the potential differences between different individuals. However, it is not known how much the data quality is impacted by the limitations of the eye-tracker, untrained participants, and the unsupervised experiment. Therefore, it is not clear what we can learn from this experiment and whether similar installations may be a way forward towards assessing huge numbers or participants, potentially revealing differences in gaze behavior. Furthermore, to maximize the impact of these experiments new automated data analysis approaches are needed to classify gaze events and to process such a large amount of data. While a large number of such algorithm exist, it is important to investigate which one of them performs the best and how that choice affects the results of data analysis and, thus, the conclusions. The other point is that the large amount of data might compensate the potentially lower quality of data. The aim of this graduation assignment is to use freely available coding tools to analyze a unique and large dataset collected with unsupervised eye-tracking experiment. Further, the data will allow for explorative analyses into the relationship of central gaze parameters and demographic variables. Data analysis approaches are inspired by different research fields including image processing and machine learning
dc.description.sponsorshipUtrecht University
dc.format.extent3216113
dc.format.mimetypeapplication/pdf
dc.language.isoen
dc.titleArtificial intelligence for the predication of demographics from unsupervised eye tracking data
dc.type.contentMaster Thesis
dc.rights.accessrightsOpen Access
dc.subject.keywordsData analysis, Eye tracking, Machine learning
dc.subject.courseuuApplied Cognitive Psychology


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record