Sie befinden Sich nicht im Netzwerk der Universität Paderborn. Der Zugriff auf elektronische Ressourcen ist gegebenenfalls nur via VPN oder Shibboleth (DFN-AAI) möglich. mehr Informationen...
Designing Researchmap: A Revolutionary Scholar Support Platform Achieved Through Human-AI Collaboration
Ist Teil von
2023 IEEE 6th International Conference on Knowledge Innovation and Invention (ICKII), 2023, p.367-371
Ort / Verlag
IEEE
Erscheinungsjahr
2023
Quelle
IEEE Xplore
Beschreibungen/Notizen
Researchmap is a researcher support platform software-as-a-service (SaaS) that has successfully achieved the semi-automated construction of a comprehensive catalog of achievements for over 340,000 Japanese researchers through the collaboration of artificial intelligence (AI) and human expertise. Researchmap utilizes APIs from primary information providers such as Scopus to compile a substantial portion of researchers' achievement lists. This curated dataset then serves as training data for AI development. The AI model is iteratively refined by providing feedback to the users (i.e.researchers), allowing them to make corrections and improvements using a user-friendly interface. As a result of these efforts, Researchmap successfully captures information from more than 340,000 researchers within the Japanese academia, accumulating over 20 million research achievements in machine-readable formats. The F1 Score of AI in recommending discovered research achievements to the respective researchers exceeds 0.94.