Difference between revisions of "Zhou et al ACM symposium on Applied Computing 2006"

From Cohen Courses
Jump to navigationJump to search
Line 9: Line 9:
 
== Summary ==
 
== Summary ==
  
−
a MEDical Information Extraction (MedIE) system that extracts and mines a variety of patient information from free-text clinical records. Their approaches are:
+
The paper presents about a MEDical Information Extraction (MedIE) system, which extracts and mines a variety of patient information from free-text clinical records.  
  
−
* A graph-based approach which uses the parsing result of link-grammar parser for relation-extraction
+
* extraction of medical terms
 +
e.g.
 +
* text classification
 +
e.g. a patient can be classified as a former smoker, a current smoker, or a non-smoker
 +
* relation extraction
  
 +
Their approaches are:
 
* An ontology-based approach for extracting medical terms of interest
 
* An ontology-based approach for extracting medical terms of interest
−
 
 
* an NLP-based feature extraction method coupled with an ID3-based decision tree for text classification
 
* an NLP-based feature extraction method coupled with an ID3-based decision tree for text classification
 +
* A graph-based approach which uses the parsing result of link-grammar parser for relation-extraction
  
 +
This approach was fairly successful
  
−
This is an early and influential [[Category::paper]] presenting an unsupervised approach to [[AddressesProblem::review classification]]. The basic ideas are:
+
However, the size of data set used is small. When more diversified writing styles are introduced into patient records, the performance may degrade. link grammar parser makes many errors.
−
 
 
−
* To use patterns of part of speech tags to pick out phrases that are likely to be meaningful and unambiguous with respect to semantic orientation (e.g. ADJ NOUN might pick out "good service" or "delicious desserts").
 
−
 
 
−
* To use [[UsesMethod::pointwise mutual information]] (PMI) to score the similarity of each phrase in a review with the two words "excellent" or "poor", and give an overall score for the polarity to each phrase based on the difference of its PMI with "excellent" to the PMI with "poor". A large corpus was used here (the Web, via queries to a search engine).
 
−
 
 
−
* To score the polarity of a review based on the total polarity of the phrases in it.  
 
−
 
 
−
This approach was fairly successful on a range of review-classification tasks: it achieved accuracy of between 65% and 85% in predicting an author-assigned "recommended" flag for Epinions ratings for eight diverse products, ranging from cars to movies. Many later writers used several key ideas from the paper, including: treating polarity prediction as a document-classification problem; classifying documents based on likely-to-be-informative phrases; and using unsupervised or semi-supervised learning methods.
 
  
 
== Related papers ==
 
== Related papers ==

Revision as of 19:17, 30 September 2010

Citation

Ciaohua Zhou et al. 2006. Approaches to Text Mining for Clinical Medical Records. In Proceedings of the 2006 ACM symposium on Applied computing, 235-239.

Online version

ACM portal

Summary

The paper presents about a MEDical Information Extraction (MedIE) system, which extracts and mines a variety of patient information from free-text clinical records.

  • extraction of medical terms

e.g.

  • text classification

e.g. a patient can be classified as a former smoker, a current smoker, or a non-smoker

  • relation extraction

Their approaches are:

  • An ontology-based approach for extracting medical terms of interest
  • an NLP-based feature extraction method coupled with an ID3-based decision tree for text classification
  • A graph-based approach which uses the parsing result of link-grammar parser for relation-extraction

This approach was fairly successful

However, the size of data set used is small. When more diversified writing styles are introduced into patient records, the performance may degrade. link grammar parser makes many errors.

Related papers

The widely cited Pang et al EMNLP 2002 paper was influenced by this paper - but considers supervised learning techniques. The choice of movie reviews as the domain was suggested by the (relatively) poor performance of Turney's method on movies.

An interesting follow-up paper is Turney and Littman, TOIS 2003 which focuses on evaluation of the technique of using PMI for predicting the semantic orientation of words.

Retrieved from "http://curtis.ml.cmu.edu/w/courses/index.php?title=Zhou_et_al_ACM_symposium_on_Applied_Computing_2006&oldid=1929"