Presentation Information
[A-12-09]Investigating the Effects of Task Characteristic Differences on AI Assistance Effectiveness in Crowdsourcing
◎△Minami Nagaoka1, Ryota Noseyama2, Motoki Bamba2, Akihito Kohiga1, Takahiro Koita2 (1. Doshisha Univ., 2. Graduate School of Doshisha Univ.)
Keywords:
Crowdsourcing
Against the background of a declining labor force caused by population aging and the diversification of work styles, crowdsourcing has attracted attention as a flexible work arrangement that can utilize diverse human resources regardless of time and place. However, because crowdsourcing has low entry barriers, workers with different levels of experience and skills tend to coexist, making quality assurance difficult. In addition, short-term contracts make continuous training difficult, creating a need for support methods that stabilize output quality regardless of workers’ abilities. Although AI-based worker support has recently attracted attention, its effectiveness is not always consistent, and the task characteristics that influence its effects remain unclear. This study focuses on higher-order cognitive activities in Bloom’s taxonomy and examines the effects of AI support through a two-factor between-subjects experiment using product reviews. In the AI support condition, thinking frameworks and examples were provided. The results showed that the effect of AI support differed by cognitive activity, suggesting the importance of designing support that fits task characteristics.