Dissertation Title: Applying Artificial Intelligence techniques to help achieve automated decision-making
Authors: Ilma Lili
- Institution: University of Tirana, Faculty of Natural Sciences, Department of Informatics
- Field of study: computing
- Publication date: 06.08.2026
- The dissertation is published in Albanian.
© Copyright: Ilma Lili
Published by the University of Tirana
Based on legal acts, regulations and policies of the UT.
👉 Click here to view the full dissertation (PDF)
Abstract: This dissertation addresses the role of Artificial Intelligence techniques in helping to implement automated decision-making systems in various application contexts. The study focuses on analyzing the relationship between the performance of intelligent models, the transparency of decisions and the integration of autonomous systems in real environments. The paper addresses the theoretical concepts of automated decision-making, probabilistic models, Machine Learning and Deep Learning techniques, as well as Explainable Artificial Intelligence (XAI) approaches. From a methodological perspective, the dissertation presents three main case studies. The first case deals with the integration of XAI methods into intelligent chatbots for the hospitality sector, with the aim of increasing the transparency and reliability of automated decisions. The second case focuses on sentiment analysis from Albanian speech, using Deep Learning models and acoustic representations such as MFCC and HuBERT in a context with limited linguistic resources. The third case proposes a framework based on the Digital Twin concept for monitoring indoor air quality and realizing automatic decision-making through IoT sensors, predictive models and adaptive control modules. The results show that Artificial Intelligence techniques can be effectively used to automate decision-making processes, increasing the accuracy, speed and predictive ability of systems along with the integration of explainability methods.
Keywords: Artificial Intelligence, automated decision-making, Machine Learning, Deep Learning, XAI (eXplainable AI), Digital Twin, sentiment analysis, IoT, HuBERT, intelligent chatbot
Abstract: This dissertation examines the role of Artificial Intelligence techniques in supporting the implementation of automated decision-making systems across different application contexts. The study focuses on analyzing the relationship between the performance of intelligent models, decision transparency, and the integration of autonomous systems into real-world environments. Within the scope of this work, the theoretical concepts of automated decision-making, probabilistic models, Machine Learning and Deep Learning techniques, as well as Explainable Artificial Intelligence (XAI) approaches are addressed. From a methodological perspective, the dissertation presents three main case studies. The first case study examines the integration of XAI methods into intelligent chatbots for the hospitality sector, with the aim of increasing the transparency and reliability of automated decisions. The second case study focuses on speech sentiment analysis in the Albanian language by using Deep Learning models and acoustic representations such as MFCC and HuBERT within a low-resource language context. The third case study proposes a framework based on the concept of the Digital Twin for indoor air quality monitoring and the implementation of automated decision-making through IoT sensors, predictive models, and adaptive control modules. The results demonstrate that Artificial Intelligence techniques can be effectively used to automate decision-making processes by improving the accuracy, speed, and predictive capabilities of the systems, together with the integration of explainability methods.
Keywords: Artificial Intelligence, automated decision-making, Machine Learning, Deep Learning, XAI (eXplainable Artificial Intelligence), Digital Twin, sentiment analysis, IoT, HuBERT, intelligent chatbot