A survey of current work in biomedical text mining

From Brede Wiki
Jump to: navigation, search
Paper (help)
A survey of current work in biomedical text mining
Authors: Aaron M. Cohen, William R. Hersh
Citation: Briefings in Bioinformatics 6 : 57-71. 2005
Database(s): PubMed (PMID/15826357)
DOI: 10.1093/bib/6.1.57.
Link(s): http://skynet.ohsu.edu/~hersh/briefings-05-cohen.pdf
Web: Bing Google Yahoo!Google PDF
Article: BASE Google Scholar PubMed
Restricted: DTU Digital Library
Other: NIF
Format: BibTeX Template from PMID
Extract: Talairach coordinates from linked PDF: CSV-formated wiki-formated

A survey of current work in biomedical text mining is a review on biomedical text mining.

They divide the task into:

  1. Named entity recognition
  2. Text classification
  3. Synonym and abbreviation extraction
  4. Relationship extraction
  5. Hypothesis generation
  6. "Integration frameworks"

[edit] Named entity recognition

Concepts mentioned are part-of-speech tagging, evaluation by precision, recall and F-score.

Approached mentioned are (mostly for gene and protein NER):

  1. AbGene by Tanabe and Wilbur based on Brill POS tagger[1]
  2. GAPSCORE by Tanabe and Wilbur, Chang et al.[2]
  3. Yapex corpus, see [3]
  4. Zhou et al. hidden Markov model.[4]
  5. GENIA corpus.[5]
  6. MEDLEE.[6]

[edit] Related papers

  1. A review of feature selection techniques in bioinformatics

[edit] References

  1. Tagging gene and protein names in biomedical text
  2. GAPSCORE: Finding gene and protein names one word at a time
  3. Protein names and how to find them
  4. Recognizing names in biomedical texts: a machine learning approach
  5. GENIA corpus - a semantically annotated corpus for bio-textmining
  6. Extracting phenotypic information from the literature via natural language processing
Personal tools