FCCC framework for conversational AI systems : designing self-regulating systems with fuzzy conversational character computing
BP2-STS
- Hundertmark, Sophie ORCID Université de Fribourg
- Fribourg, Switzerland, 2026
1 ressource en ligne (XI, 219 pages) ; 1 fichier pdf
PhD: Université de Fribourg (Suisse), 30.03.2026
English
This dissertation presents the Fuzzy Conversational Character Computing (FCCC) framework, a new conceptual framework for designing conversational AI systems that act as self-regulating, adaptive and ethically aligned. It builds upon four theoretical foundations: systems theory (von Bertalanffy, 1949) for self-regulation, fuzzy logic (Zadeh, 1965) to manage linguistic ambiguity, conversational theory (Pask, 1965; Pangaro, 2010) to model dynamic human–machine feedback loops, and Character Computing (El Bolock et al., 2020) to integrate personality, affective state and behaviour.
Using Design Science Research (Möller et al., 2020), the thesis combines theoretical synthesis, empirical investigation and artefact evaluation. Four independent use cases yielded seven design requirements for the FCCC artefact. The findings reveal how conversational AI can balance user expectations, company objectives and ethical responsibility while ensuring sustainable interaction quality.
The FCCC Framework connects conversational loops with multidimensional fuzzy classifications (Zumstein, 2007, 2026) and Character Computing, enabling continuous learning from user input and context. Its inner part is mathematically formulated and implemented as a prototype chatbot integrating fuzzy personality vectors with large language models, demonstrating how emotional inputs translate into adaptive responses that dynamically adjust tone, style and empathy.
Focus groups and expert interviews confirm the framework's organizational feasibility and ethical relevance within Life Engineering. By prioritizing transparency, user well-being and responsible automation, the FCCC Framework contributes to "life agents" that support human flourishing, uniting psychology, computer science and business informatics into a sustainable, empathetic and explainable AI architecture.
Using Design Science Research (Möller et al., 2020), the thesis combines theoretical synthesis, empirical investigation and artefact evaluation. Four independent use cases yielded seven design requirements for the FCCC artefact. The findings reveal how conversational AI can balance user expectations, company objectives and ethical responsibility while ensuring sustainable interaction quality.
The FCCC Framework connects conversational loops with multidimensional fuzzy classifications (Zumstein, 2007, 2026) and Character Computing, enabling continuous learning from user input and context. Its inner part is mathematically formulated and implemented as a prototype chatbot integrating fuzzy personality vectors with large language models, demonstrating how emotional inputs translate into adaptive responses that dynamically adjust tone, style and empathy.
Focus groups and expert interviews confirm the framework's organizational feasibility and ethical relevance within Life Engineering. By prioritizing transparency, user well-being and responsible automation, the FCCC Framework contributes to "life agents" that support human flourishing, uniting psychology, computer science and business informatics into a sustainable, empathetic and explainable AI architecture.
- Faculty
- Faculté des sciences économiques et sociales et du management
- Language
-
- English
- Classification
- Medicine, Technology, Engineering
- Notes
-
- Bibliographie
- License
-
CC BY
- Open access status
- diamond
- Identifiers
-
- SWISSCOVERY 991172966513205501
- DOI 10.51363/unifr.eth.2026.034
- URN urn:nbn:ch:rero-002-123912
- Persistent URL
- https://folia.unifr.ch/unifr/documents/335279
Statistics
Document views: 3
File downloads:
- phd_Hundertmarks.pdf: 1