This paper deals with the problem of recognizing and extracting acronymdefinition pairs in Swedish medical texts. This project applies a rule-based method to solve the acronym recognition task and compares and evaluates the results of different machine learning algorithms on the same task. The method proposed is based on the approach that acronym-definition pairs follow a set of patterns and other regularities that can be usefully applied for the acronym identification task. Supervised machine learning was applied to monitor the performance of the rule-based method, using Memory Based Learning (MBL). The rule-based algorithm was evaluated on a hand tagged acronym corpus and performance was measured using standard measures recall, precision and f-score. The results show that performance could further improve by increasing the training set and modifying the input settings for the machine learning algorithms. An analysis of the errors produced indicates that further improvement of the rule-based method requires the use of syntactic information and textual pre-processing.
[1]
Dimitrios Kokkinakis.
Collection, Encoding and Linguistic Processing of a Swedish Medical Corpus - The MEDLEX Experience
,
2006,
LREC.
[2]
Marti A. Hearst,et al.
A Simple Algorithm for Identifying Abbreviation Definitions in Biomedical Text
,
2002,
Pacific Symposium on Biocomputing.
[3]
Youngja Park,et al.
Hybrid Text Mining for Finding Abbreviations and their Definitions
,
2001,
EMNLP.
[4]
Peter D. Turney,et al.
A Supervised Learning Approach to Acronym Identification
,
2005,
Canadian AI.
[5]
J. Meigs,et al.
WHO Technical Report
,
1954,
The Yale Journal of Biology and Medicine.
[6]
Stuart Yeates,et al.
Automatic Extraction of Acronyms from Text
,
1999,
New Zealand Computer Science Research Students' Conference.
[7]
Thomas G. Szymanski,et al.
A fast algorithm for computing longest common subsequences
,
1977,
CACM.
[8]
Paul Ogilvie,et al.
Acrophile: an automated acronym extractor and server
,
2000,
DL '00.