BrainKnow -- Extracting, Linking, and Associating Neuroscience Knowledge

The vast accumulation of neuroscience knowledge presents a challenge for researchers to timely and accurately locate the specific information they require. Constructing a knowledge engine that automatically extracts and organizes information from academic papers can provide researchers with timely and accurate informational services. We present the Brain Knowledge Engine (BrainKnow), which extracts and integrates neuroscience knowledge from published papers from PubMed. BrainKnow comprises a substantial repository, containing 3,626,931 relations spanning a broad spectrum of 37,011 neuroscience concepts extracted from 1817744 articles. The relations in BrainKnow can be accessed and navigated through a user-friendly web interface. Additionally, BrainKnow employs graph network algorithms for the recommendation and visualization of knowledge. BrainKnow is capable of automatic real-time updates. BrainKnow represents the first neuroscience knowledge graph that not only integrates knowledge in-depth but also facilitates fully automated updates.