• FlexiTerm is an open-source software tool for automatic recognition of multi-word terms.
  • FlexiTerm uses a range of methods to neutralise the main sources of term variation including acronyms.
  • FlexiTerm has been evaluated on five domain-specific corpora (see table below).
  • FlexiTerm is robust enough for less formally structured texts such as those found in patient blogs or medical notes.

Data set Topic Document type Search terms Sample output
1 molecular biology PubMed abstract human, blood cell, transcription factor output 1
2 COPD PubMed abstract "pulmonary disease, chronic obstructive" [MeSH Terms] output 2
3 COPD blog post COPD, chronic obstructive {pulmonary | lung | airways | respiratory} disease, bronchitis, emphysema output 3
4 obesity, diabetes hospital discharge summary N/A output 4
5 knee MRI report N/A output 5


  • FlexiTerm source files written in Java (distributed under the FreeBSD licence). This is the latest version described in the 2018 paper.
  • MinorThird can be used to annotate term occurrences recognised by FlexiTerm. You can see an example here.
  • SQLite database is used by FlexiTerm to manage intermediary results.
  • SQLite Manager can be used to view the SQLite database.