Abstract
Crowdsourcing contests have emerged as a popular approach for organisations to leverage collective intelligence and tap into the expertise of a diverse crowd. Knowledge sharing in such contests is paramount in promoting individual or team participation. This study employs negative binomial regression analysis to explore the impact of dynamic knowledge-sharing features - knowledge volume, knowledge expansion, knowledge innovation, and knowledge popularity - on team participation, using data from 211 crowdsourcing contests hosted on Kaggle. The findings offer significant implications for researchers and practitioners regarding effective knowledge management within crowdsourcing platforms. Encouraging knowledge sharing and collaboration and promoting innovative practices can significantly boost the attraction of more participating teams while amplifying the volume of submissions. Additionally, this study underscores the importance of implementing mechanisms to mitigate knowledge overload and ensure diversity of ideas, thereby sustaining engagement and promoting more unique submissions.