Assessing the Impact of Machine Learning-Based Personalization on User Experience in E-Commerce Websites

Thesis Abstract

The rapid proliferation of e-commerce platforms has underscored the importance of personalized user experiences in enhancing customer satisfaction, engagement, and loyalty. This study addresses the growing reliance on machine learning algorithms to dynamically tailor content and recommendations, and critically assesses their impact on user experience within e-commerce environments, recognizing a research gap concerning empirical evidence on behavioral and perceptual outcomes. The primary aim of this research is to evaluate how machine learning-based personalization influences user satisfaction, perceived usability, and purchase intention on major e-commerce websites. Specific objectives include (1) identifying the predominant machine learning techniques employed for personalization, (2) measuring user perceptions of personalized recommendations, (3) examining the relationship between personalization and key user experience metrics, and (4) providing empirically grounded recommendations for optimizing machine learning applications in e-commerce settings. Employing a quantitative research design, the study utilizes a cross-sectional survey approach combined with experimental testing to collect comprehensive data. The population comprises active users of five leading e-commerce platforms operating within the retail sector, totaling approximately 10,000 individuals. A stratified random sampling technique extracts a sample size of 384 respondents, ensuring representation across demographic groups. Data collection instruments include a structured questionnaire adapted from validated scales in user experience and e-commerce research, alongside behavioral data obtained from platform analytics to measure interaction patterns and recommendation engagement. Prior to data collection, the instruments undergo validation through expert review and pilot testing to establish validity and reliability, with Cronbach’s alpha coefficients exceeding 0.8 for all scales. Data analysis employs multiple linear regression to examine the relationship between the intensity of personalization (independent variable) and user experience measures (dependent variables), complemented by ANOVA to compare differences across demographic groups. Additionally, thematic analysis of open-ended survey responses provides qualitative insights into user perceptions and attitudes regarding personalized recommendations. The theoretical framework integrates the Technology Acceptance Model (TAM) to elucidate perceived usefulness and ease of use, and the Expectation Confirmation Theory (ECT) to contextualize satisfaction levels. The study also develops a conceptual model illustrating the pathways through which machine learning personalization influences user experience. Expected findings reveal that effective personalization significantly enhances perceived usability, satisfaction, and purchase intention, with machine learning techniques such as collaborative filtering and deep learning offering superior personalization outcomes. The research anticipates identifying demographic and behavioral moderators affecting the strength of these relationships. The study’s contribution to knowledge includes providing empirical validation of theoretical models in the context of machine learning-driven e-commerce personalization, and offering evidence-based insights for practitioners seeking to optimize recommender system deployment. The main conclusion emphasizes that tailored algorithms positively affect user experience, but their effectiveness depends on implementation quality and user perceptions. Based on these findings, recommendations include adopting transparent personalization strategies, employing hybrid recommendation techniques, and continually monitoring user feedback to adapt algorithms accordingly. The study advocates for further longitudinal research to explore long-term impacts of personalization and the integration of emerging machine learning methods to sustain user engagement and trust in e-commerce platforms. Overall, this research advances understanding of the nuanced dynamics between machine learning personalization and user experience, providing a foundation for more user-centric and technologically innovative e-commerce practices.