Towards automated analysis of eye tracking studies using the workflow technology

Abstract

Eye tracking studies have become one means to evaluate user behavior. For example, eye tracking is used in marketing research, psychology, human-computer interaction, or visualization. Analyzing eye movement data can help to find where on a stimulus and at what areas of a stimulus a participant focused on. Eye tracking can be used besides classical benchmark metrics such as completion times and accuracy rates of correctly given answers. However, evaluating eye movement data is a time consuming task, as a large amount of data has to be analyzed. Typically, multiple software systems have to be used, each having a different format of the data. This leads to tedious work, as the analyst has to reformat the data and learn how to use different software systems. Therefore, we suggest to analyze eye movement data using workflow technology. This allows to automatically analyze the data, and have a reproducible result. The analyst is not concerned with the technicalities and can rather focus on the interpretation of the analyzed data. Allowing an automatic evaluation using the workflow technology, requires that the analysis functionality is available via web services. In this paper, we contribute a workflow system based on web service operations. The web service offers functionality of an eye tracking analysis framework for the automatic analysis of eye movement data. The workflow system can be used to model and execute different types of eye tracking evaluations.