Competitive gamification in crowdsourcing-based contextual-aware recommender systems

Area of Science:

• Human-Computer Interaction
• Information Systems
• Epidemiology

Background:

• Crowdsourcing-based context-aware recommender systems (CARS) are vital for contactless decision-making during pandemics like COVID-19.
• Understanding the impact of gamification on user engagement in crowdsourcing for real-time data collection is crucial.

Purpose of the Study:

• To investigate the effectiveness of crowdsourcing-based CARS in supporting user decisions during epidemics.
• To analyze how different game designs influence user participation in crowdsourcing tasks for restaurant recommendations.

Main Methods:

• Development of a crowdsourcing-based CARS for restaurant recommendations during COVID-19.
• A two-week field study with 68 users across four conditions: control, self-competitive, social-competitive, and mixed gamification.
• System provided real-time recommendations based on restaurant epidemic status.

Main Results:

• Crowdsourcing is feasible for collecting real-time information for recommendations during epidemics.
• Mixed competitive gamification increased engagement for both high- and low-performance users.
• Self-competitive game design motivated users to undertake a broader range of tasks.

Conclusions:

• Crowdsourcing-based CARS are viable for providing timely, context-aware recommendations during health crises.
• Gamification strategies, particularly mixed competition and self-competition, can significantly enhance user engagement and task diversity in crowdsourcing systems.