Building an Integrated Scientific Knowledge Graph for Neuroscience: Enhancing Research Impact Analysis and Optimizing Data Curation

The Neuroscience pilot of the EU-funded SciLake project seeks to tackle the complexities of fragmented and disorganized open research data. This initiative has united partners with expertise in knowledge management with curators from EBRAINS (RRID: SCR_019260), an open research infrastructure focused on brain-related studies. The innovative Scientific Knowledge Graph (SKG) (https://bip.imsi.athenarc.gr/search/neuroscience-pilot) created through this pilot serves as a cohesive framework for data integration while enabling the use of value-added services provided by SciLake. The developed SKG combines curated datasets from EBRAINS with extensive neuroscience research products from the OpenAIRE neuroscience gateway (https://neuroscience.openaire.eu/). The SKG incorporates controlled terms from openMINDS (RRID: SCR_023173) as nodes and utilizes an entity recognition model to classify and map these terms in relevant texts, enriching the connections among research products. Additionally, the SciLake services deploy language models to extract mentions of research artifacts (RAs) and enhance the interlinking of research products. Fully integrated with impact-driven discovery and reproducibility tools and leveraging the analytical capabilities of SciLake, the SKG serves as a powerful resource for contextualizing curation efforts and exploring research trends, ultimately optimizing curation processes in the neuroscientific landscape.