Seeking Abbreviations From MEDLINE Jeffrey T. Chang Hinrich Schütze Russ B. Altman Presented by: Bo Han
Challenge n Huge –Contains 12 million citations back to 1966 n Growing –400,000 citations per year n Common and Uncontrolled Use of Abbreviates –Authors create abbreviations in different ways
Example: Abb. Def. n VDR vitamin D receptor n PTU propylthiouracil n JNK c-Jun N-terminal kinase n IFN interferon n ATL adult T-cell leukemia n Beta-EP beta-endorphin
Previous Methods n Match Initial letters (Acronyms) filter out some common words Office of Nuclear Waste Isolation (ONWR) n Use Heuristics Favor matches on the first letter or syllable boundaries, Upper case Letters difficulty: finding optimal weights n Use Lexicon Rules
Motivation An automatic method for detecting abbreviations first step for understanding biological literatures Rather than text mining skills, using sequence alignment to find candidates Classification algorithm is used to confirm candidates
Method Scanning text for possible abbreviations Aligning the candidates Converting the abbreviations and alignments into a feature vector Classify by a machine learning method
Finding Candidates n Pattern long form (abbreviation) n Assumption: ignore abb. longer than two words n Within Parentheses, stop at a comma/semicolon n 3N for long form
Aligning Abbreviations n Longest Common Substring(LCS) n Seq Alignment: Maximize # of matched letters long form --- abbreviation O(NM)
Aligning Abbreviations n Antioxidant response element A R E
Computing Features in an alignment n Rather using scoring matrix, they use feature vector n Choosing Features Lower/Upper Case Beginning of Word 5.54 End of Word Syllable Boundary 2.08 After Aligned Letter 1.50 Letters Aligned 3.67 Words Skipped Aligned Letters per Word 0.70 Constant -9.70
Scoring Alignment n Supervised Learning n Training set=1000 random candidates n Binary Logistic Regression Classifier P: probability of seeing an abbreviations X: feature vector W: weight vector Finding w which maximized the difference between positive and negative examples
Evaluation n A gold standard corpus n Recall (84%) Ten correct not retrieved out of 50 correct docs 40/50=80% n Precision(81%) 65 out of 100 docs are relevant 65/100=65%
Test it on-line
Disadvantage n Strict form n Without considering the context n Some grammar rule is ignored in feature vector ubMed&list_uids= &dopt=Abstract
Conclusions A novel algorithm for finding abbreviation Combing sequence alignment and machine learning Further work is expected to improve performance
Reference Chang JT, Schütze H, and Altman RB (2002). Creating an Online Dictionary of Abbreviations from MEDLINE. The Journal of the American Medical Informatics Association. 9(6): Acronym finder. Iliopoulos I, Enright A, Ouzounis C. Textquest: document clustering of medline abstracts for concept discovery in molecular biology. Pac Symp Biocomput 2001;