What we do

The BioMeXT group specializes in Information Extraction from biomedically relevant textual sources, such as the scientific literature, clinical records, or social media. We focus in particular on the extraction of domain-specific entities (such as genes, drugs, diseases), and their semantic relationships (e.g. gene-disease associations). Our tools are often evaluated through participation in community-run evaluation challenges (e.g. BioCreAtIvE). We also provide an environment for Assisted Curation (ODIN), which is used in the curation pipeline of the RegulonDB database, in an NIH-funded project.

Find out about the Group’s activities

Main publications 2019

  • Furrer L et al.
    OGER++: hybrid multi-type entity recognition
    Journal of Cheminformatics, doi: 10.1186/s13321-018-0326-3
  • Sheikhalishahis S et al.
    Natural Language Processing of Clinical Notes on Chronic Diseases: Systematic Review
    JMIR Med Inform, doi: 10.2196/12239
  • Vishnyakova D et al.
    A new approach and gold standard toward author disambiguation in MEDLINE
    J Am Med Inform Assoc, doi: 10.1093/jamia/ocz028

Find out more about the Group’s activities


University lugano

Fabio Rinaldi
BioMeXT: Biomedical Information Extraction
Dalle Molle Institute for Artificial Intelligence Research (IDSIA)
Group Webpage

Domain(s) of activity:

  • Text mining and machine learning
  • APIs
  • Electronic health record
  • Machine learning
  • Ontology
  • Text mining

Domain(s) of application:

  • Basic research
  • Medicine and health