Biomedical literature

Semantic Indexing & Question answering​

  • ASiS Open Data Graph 
      • an open-source framework for the semantic retrieval and integration of biomedical knowledge
      • Topic-specific, for any topic of interest
      • From both unstructured and structured resources
      • Fully automated construction and incremental update
      • Graph representation under an intuitive semantic schema 

 

  • Up-to-date topic-specific Knowledge Graphs 
      • Easy access and browsing of knowledge for biomedical researchers

Open Knowledge Integration

Biomedical literature is currently indexed with topic labels from the Medical Subject Headings (MeSH) thesaurus

About 35% of the MeSH labels cover multiple domain concepts

  • Νo annotations are available at this fine-grained level of detail
  • BeyondMeSH: A method for fine-grained semantic indexing;
    Weakly supervised learning based on concept occurrence

MeSH Extension: More than 300 new descriptors annually

  • MeSH conceptual provenance codes: A framework for identifying and characterizing new descriptors based on their relation to their predecessors.