| Abstract [eng] |
The primary objective of this master's thesis is to explore the dynamics and effectiveness of personalization and contextual factors in AI based recommender systems. The thesis is structured into three main sections: review of existing literature, an empirical research study, and the presentation of research results. The literature review delves into various types of recommendation systems, discusses prevalent challenges such as data sparsity, privacy concerns, scalability, and latency, alongside the potential benefits like enhanced user engagement and increased sales conversion rates. In this part, the main contextual and personalization factors used in recommender systems are identified. This section also highlights the significance of personalization and contextual factors in enhancing the effectiveness of these systems. An empirical study was conducted to investigate the impact of personalization factors (demographic information, social connections, psychophysiological aspects) and contextual factors (season, geographic location, day of the week, time of day, weather conditions, noise level) on user satisfaction and trust in AI recommender systems. The study involved 311 respondents, utilizing questionnaires to gather data on their experiences and perceptions. Statistical tools, including correlation and regression analysis, were employed to analyze the data. The findings revealed nuanced insights into user preferences and the varied significance of different personalization and contextual elements in recommendation systems. The thesis concludes with a summary of key findings: personalized recommendation systems are more trusted and preferred by users, with specific personalization and contextual factors being critical for enhancing user satisfaction. Recommendations are provided for future research, emphasizing the need to explore unaccounted variables that might influence user perceptions of personalized recommendations. For businesses, the thesis suggests focusing on personalization and contextual factors to enhance user engagement and satisfaction with AI-driven recommender systems. |
