Analyzing Text-Based User Feedback in e-Government Services Using Topic Models

In providing e-Government services, it is necessary to conduct a user-centric evaluation periodically, to improve the usability, accessibility, and effectiveness of e-Government services for citizens. In a qualitative evaluation, gathering user´s opinions and recommendations to the e-Government services are commonly used to get a real feedback from the citizens. However, such the user feedback commonly is containing a big text data from diversity of the citizens. It is necessary to do a text mining to assist an analyzing to find out quickly the point of what the citizens recommended. This study proposes an application of Topic Models to do analysis of the big text datasets of the user´s feedback. Based on the qualitative analyses approach, the study analyzes user´s opinion and recommendation to the e-Government services. The paper also describe designing evaluation sheet, data collection and analysis strategies of a case study in e-Government portals evaluation.