Assessment of the University Digital Ecosystem through Intelligent Analysis of Open-Ended Survey Responses

Abstract

The use of open-ended questions in surveys for conducting research in various fields of activity allows you to obtain detailed and in-depth information that can be used to make important decisions. However, automatic processing of open questions is a complex task that requires an integrated approach. This study proposes a methodology based on sentiment analysis, topic modeling, and text networks to extract useful data on satisfaction with the university's digital ecosystem and the current situation. The methodology allows you to get a more complete picture of the level of satisfaction of students, lecturers and other participants in the university's digital ecosystem, as well as identify problematic issues and offer specific recommendations for their improvement. The main contribution of the article is the use of algorithms for modeling topics and network structures to complement the results of the analysis of the tone of the text. The methodology includes Data Gathering, Text Processing, Text Mining and Recommendations. The proposed methodology is applicable not only to the educational field, but can also be replicated for automatic processing of unstructured data in other areas.