Begin: Tuesday 30. of July 2019 (6:15 PM)
Location: Warburger Str. 100, Lecture Room O2
User satisfaction, or the overall cognitive and affective evaluation of an individual’s experiences with an IT artifact, has been a subject of interest for IS researchers since the inception of the field. Since cognitive and affective evaluations do not stay the same over prolonged periods of time, and today’s IT artifacts are subject to continuous releases and bug fixes, we are interested in unearthing the fluctuations in user satisfaction and its antecedents. By applying dynamic topic modeling and sentiment analysis based on a pre-trained recurrent neural network on more than 60,000 mobile app reviews, this paper tracks the levels of user satisfaction with a mobile app over a 6-year period and shows the dynamic interplay of variables during that time frame. Furthermore, the paper provides a new perspective on data collection and analysis that can supplement existing methods in their ambition to shed light on the nature and antecedents of evaluative constructs and to decipher the strength of their established relationships over time.