THE IMPACT OF ARTIFICIAL INTELLIGENCE-BASED RECOMMENDATION SYSTEMS ON CONSUMER PURCHASE DECISIONS IN E-COMMERCE

Abstract

This study investigates the impact of artificial intelligence-based recommendation systems on consumer purchase intention within e-commerce platforms, focusing on the multidimensional influence of recommendation accuracy, system personalization, user interface quality, trust and privacy, and perceived value in contemporary digital marketplace contexts. The exponential growth of e-commerce platforms has created unprecedented opportunities for AI-driven recommendation systems to influence consumer behavior, yet comprehensive understanding of how multiple system dimensions collectively shape purchase decisions remains limited in existing literature. This research addresses this gap by examining the simultaneous effects of five critical AI recommendation system characteristics on consumer purchase intentions through a quantitative correlational approach. Data were collected from 100 active e-commerce users through a structured questionnaire employing 5-point Likert scales, with analysis conducted using SPSS version 26. Rigorous methodological procedures included validity testing with correlation coefficients exceeding r > 0.195, reliability assessment with Cronbach's Alpha values above 0.70 for all constructs, and comprehensive classical assumption testing to ensure statistical validity. Multiple regression analysis revealed significant positive effects for all examined variables: System Personalization demonstrated the strongest influence (t = 3.921, p < 0.001), followed by Trust and Privacy (t = 3.456, p = 0.001), Recommendation Accuracy (t = 2.847, p = 0.006), Perceived Value (t = 2.389, p = 0.019), and User Interface Quality (t = 2.234, p = 0.028). The overall model achieved statistical significance (F = 47.863 > F-table = 2.31, p < 0.001) with substantial explanatory power (R² = 0.718), indicating that 71.8% of purchase intention variance is explained by these five dimensions collectively. This research contributes theoretically to social commerce literature by validating a comprehensive framework that integrates technological and psychological factors influencing AI-mediated consumer behavior, while providing practical insights for e-commerce platforms seeking to optimize recommendation system effectiveness through balanced attention to personalization capabilities, trust-building measures, and user experience design.