Enhancing the Explainability of Digital Ethics Compass Through Fuzzy Logic-Based Linguistic Summaries

Supervisors: Narek Andreasyan, Edy Portmann

Contact person: Narek Andreasyan

Student: Egzont Demiri

Project status: Finished

Year: 2025

Project Abstract

This master’s thesis presents a new way to evaluate digital ethical issues using fuzzylogic based linguistic summarization. A prototype developed for this study turns complex data into simple natural language statements, making complex information explainable and easy for everyone to understand. The work underlines the importance of addressing ethical challenges in our digital world, especially those related to e-voting systems. Instead of just giving yes or no answers, this approach deals with uncertainties and partial truths to offer explainable insights. This helps both experts and everyday users comprehend the results and see how these methods can mitigate risks in digital systems. Existing methods and current applications in digital ethics assessments were explored and discussed. The approach was evaluated using a focus group and a structured questionnaire, which provided both quantitative and qualitative feedback during a demonstration session. This evaluation confirmed that users found the system useful, easy to use, and effective in making complex ethical information more explainable. Additionally, participants offered suggestions for further refinements regarding usability and for extending the linguistic summarization capabilities. Keywords: Fuzzy Logic; Linguistic

Keywords: Fuzzy Logic; Linguistic summarization; Explainability; Digital Ethics Compass; Interpretability; Transparency; Digital Services; Ethical Decision-Making; E-Voting

Document: Not yet available