A Data-Driven Expectation Prediction Framework Based on Social Exchange Theory

Area of Science:

• Data Science
• Human-Computer Interaction
• Consumer Behavior

Background:

• The proliferation of online platforms generates vast user data, often underutilized for design.
• Data-driven approaches in requirement engineering are crucial for leveraging this data.
• Social exchange theory provides a lens to understand user expectations in consumption.

Purpose of the Study:

• To propose a data-driven expectation prediction framework using social exchange theory.
• To analyze user expectations and predict design improvements for consumption platforms.
• To assist designers in making informed design enhancements.

Main Methods:

• Classified consumption exchange elements into seven categories based on social exchange resources.
• Combined word frequency statistics and scale surveys for user-generated data analysis.
• Utilized mathematical expectation formulas to predict user expectations and distinguish explicit/implicit needs.

Main Results:

• Developed a framework to predict user expectations and identify design improvement opportunities.
• Demonstrated the framework's feasibility through a case study on a service system improvement.
• Provided a method to derive reliable design improvement plans by calculating mathematical expectation.

Conclusions:

• The proposed framework offers valuable insights for data mining in consumption comments.
• Data-driven expectation prediction can significantly aid designers in enhancing user experience.
• This study bridges the gap between big data and practical design application in online consumption.