Data-Driven Consumer Preference Prediction for Product Customization Using Machine Learning and Crowdsourcing

Aug 5, 2026

Upload by Oliver

Product customization is an increasingly important paradigm shaped by manufacturing capabilities and demand for personalized goods. Implementing effective customization remains costly and complex, making decision support essential for objective strategy guidance. The thesis identifies limitations in existing decision support methodologies, including insufficient comprehensiveness, structure, and automation. It proposes a data-driven consumer preference prediction approach using 307 crowdsourced preferences and a machine-learning clustering model, achieving 70% validation accuracy with room to improve via larger datasets.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages90

Reading Time3 h 47 min

AccessFree to download

Abstract

Acknowledgements

Table of Contents

List of Tables

List of Figures

List of Symbols, Abbreviations or Nomenclature

Chapter 1 Introduction

The Customization Process

Structure of Thesis

Chapter 2 Literature Review

Classification of Customization Strategies

Key Factors in Customization Success

Decision Support Tools in Product Customization

Research Gaps

Data-Driven Consumer Preference Prediction for Product Customization Using Machine Learning and Crowdsourcing by Carter Powell A Thesis presented to The University of Guelph In partial fulfilment of requirements for the degree of Master of Applied Science in Engineering Guelph, Ontario, Canada © Carter Powell, September, 2023 ABSTRACT DATA-DRIVEN CONSUMER PREFERENCE PREDICTION FOR PRODUCT CUSTOMIZATION USING MACHINE LEARNING AND CROWDSOURCING Carter Powell University of Guelph, 2023 Advisor(s): Dr. Sheng Yang Product customization has become an increasingly popular paradigm driven by advances in manufacturing technology and demand for more personalized products. The customization process has many challenges and can be costly and complex for firms to implement. Decision support tools can help to guide firms through the customization process and offer objective recommendations on how customization strategies can be implemented. Through an extensive literary review, several shortfalls of current decision support methodologies have been found; namely that they lack comprehensiveness, structure, and automation. A data-driven approach to predicting consumer preferences for decision support is presented using 307 crowdsourced consumer preferences and a machine learning clustering model. Using a validation study, the model provides a 70% accuracy in the prediction of consumer preferences with opportunities for improvement with a larger data set. This method offers a novel approach to decision support for product customization that addresses shortfalls of other recommender systems. ACKNOWLEDGEMENTS I would like to thank Dr. Yang for all his help, guidance, and the many opportunities he has provided me with during the duration of this project. I would also like to thank my advisory committee member Dr. Lei for her guidance and support. I extend a special thank you to my family for their support and encouragement throughout my academic career. The study was supported by the funding from the Natural Science and Engineering Research Council of Canada (NSERC) Discovery Grant (RGPIN-2022-03448) . This study was approved by the University of Guelph Research Ethics Board. REB Number: 22- 12-006 TABLE OF CONTENTS Abstract ............................................................................................................................. ii Acknowledgements ............................................................................................................ iii Table of Contents ..............................................................................................................iv List of Tables ...................................................................................................................vii List of Figures ................................................................................................................. viii List of Symbols, Abbreviations or Nomenclature ..................................................................ix Chapter 1 Introduction........................................................................................................ 1 1.1 The Customization Process ........................................................................................ 2 1.2 Structure of Thesis .................................................................................................... 4 Chapter 2 Literature Review ............................................................................................... 6 2.1 Classification of Customization Strategies ................................................................... 6 2.1.1 Craft Customization ............................................................................................ 8 2.1.2 Modular Customization ....................................................................................... 9 2.1.3 Open Architecture Customization ........

FAQs

What problem does the thesis address in product customization?

Product customization is costly and complex to implement, and firms need decision support tools to choose effective customization strategies and guide execution objectively.

What data and modeling method are used for consumer preference prediction?

The approach uses 307 crowdsourced consumer preferences and a machine-learning clustering model to predict consumer preferences for customization decisions.

How accurate is the proposed prediction model?

A validation study reports 70% accuracy in predicting consumer preferences, and the thesis notes potential improvements with a larger dataset.

Aug 5, 2026

Upload by Oliver