Can AI personalise project recommendations to citizen scientists to increase their productivity and engagement?
This experiment will compare two types of intelligent recommendation algorithms that make automatic predictions about users' interests based on their past interactions with SciStarter, a portal to 3,000 global citizen science projects. The experiment will test whether these algorithms will make it easier for participants to discover the projects most suitable for them. It will also explore whether this ultimately leads to increased engagement, longer retention, improved contributions, and better quality data. It will also explore how it could increase participation by communities who don’t typically engage with citizen science projects.
Citizen scientists advance research in fields such as astronomy or zoology by sharing and analysing data, but it remains a fundamental challenge for citizen science to attract and retain enough participants to ensure the achievement of project goals. The way in which current search engines on citizen science platforms work makes it often difficult for individuals to find projects on the platform that match their interests and capabilities.
The findings of this experiment will contribute to understanding about how to improve matching algorithms to increase citizen engagement and retention rates on collective intelligence platforms. This is relevant not just for the citizen science community, but for any collective intelligence approach that involves crowdsourcing or matching individuals to specific tasks or projects.
Follow SciStarter on Twitter and stay up to date: @SciStarter