Challenges and Practical Approaches with Word Sense Disambiguation of Acronyms and Abbreviations in the Clinical Domain
2
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
Objectives: Although acronyms and abbreviations in clinical text are used widely on a daily basis, relatively little researchhas focused upon word sense disambiguation (WSD) of acronyms and abbreviations in the healthcare domain. Since clinicalnotes have distinctive characteristics, it is unclear whether techniques effective for acronym and abbreviation WSD frombiomedical literature are sufficient. Methods: The authors discuss feature selection for automated techniques and challengeswith WSD of acronyms and abbreviations in the clinical domain. Results: There are significant challenges associated withthe informal nature of clinical text, such as typographical errors and incomplete sentences; difficulty with insufficient clinicalresources, such as clinical sense inventories; and obstacles with privacy and security for conducting research with clinicaltext. Although we anticipated that using sophisticated techniques, such as biomedical terminologies, semantic types, partof-speech, and language modeling, would be needed for feature selection with automated machine learning approaches, wefound instead that simple techniques, such as bag-of-words, were quite effective in many cases. Factors, such as majoritysense prevalence and the degree of separateness between sense meanings, were also important considerations. Conclusions:The first lesson is that a comprehensive understanding of the unique characteristics of clinical text is important for automaticacronym and abbreviation WSD. The second lesson learned is that investigators may find that using simple approaches is aneffective starting point for these tasks. Finally, similar to other WSD tasks, an understanding of baseline majority sense ratesand separateness between senses is important. Further studies and practical solutions are needed to better address these issues.
Abstract
Objectives: Although acronyms and abbreviations in clinical text are used widely on a daily basis, relatively little researchhas focused upon word sense disambiguation (WSD) of acronyms and abbreviations in the healthcare domain. Since clinicalnotes have distinctive characteristics, it is unclear whether techniques effective for acronym and abbreviation WSD frombiomedical literature are sufficient. Methods: The authors discuss feature selection for automated techniques and challengeswith WSD of acronyms and abbreviations in the clinical domain. Results: There are significant challenges associated withthe informal nature of clinical text, such as typographical errors and incomplete sentences; difficulty with insufficient clinicalresources, such as clinical sense inventories; and obstacles with privacy and security for conducting research with clinicaltext. Although we anticipated that using sophisticated techniques, such as biomedical terminologies, semantic types, partof-speech, and language modeling, would be needed for feature selection with automated machine learning approaches, wefound instead that simple techniques, such as bag-of-words, were quite effective in many cases. Factors, such as majoritysense prevalence and the degree of separateness between sense meanings, were also important considerations. Conclusions:The first lesson is that a comprehensive understanding of the unique characteristics of clinical text is important for automaticacronym and abbreviation WSD. The second lesson learned is that investigators may find that using simple approaches is aneffective starting point for these tasks. Finally, similar to other WSD tasks, an understanding of baseline majority sense ratesand separateness between senses is important. Further studies and practical solutions are needed to better address these issues.
참고문헌 (30)
- Pakhomov S., "Semi-supervised Maximum Entropy based approach to acronym and abbreviation normalization in medical texts", Proceedings of the 40th Annual Meeting on Association for Computational Linguistics (ACL2002);, pp.160~167 (2002)
- Stetson PD, "The sublanguage of cross-coverage", Proc AMIA Symp, 2002, pp.742~746 (2002)
- Pakhomov S, "Abbreviation and acronym disambiguation in clinical discourse", AMIA Annu Symp Proc, pp.589~593 (2005)
- Xu H, "Machine learning and word sense disambiguation in the biomedical domain: design and evaluation issues", BMC Bioinformatics, 5(7), pp.334~ (2006)
- Xu H, "A study of abbreviations in clinical notes", AMIA Annu Symp Proc, 2007, pp.821~825 (2007)
- Kuhn IF, "Abbreviations and acronyms in healthcare : when shorter isn't sweeter", Pediatr Nurs, 33, pp.392~398 (2007)
- Walsh KE, "Medical abbreviations : writing little and communicating less", Arch Dis Child, 93(10), pp.816~817 (2008)
- Hunt DR, "Advances in patient safety: from research to implementation (Volume 2: Concepts and Methodology)" (2005), Agency for Healthcare Research and Quality
- Fan JW, "Word sense disambiguation via semantic type classification", AMIA Annu Symp Proc, 2008, pp.177~181 (2008)
- Friedman C, "Evaluating the UMLS as a source of lexical knowledge for medical language processing", Proc AMIA Symp, 2001, pp.189~193 (2001)
- Schuemie MJ, "Word sense disambiguation in the biomedical domain : an overview", J Comput Biol, 12, pp.554~565 (2005)
- Kaplan A, "An experimental study of ambiguity and context", Mech Transl, 2(2), pp.39~46 (1950)
- Choueka Y, "Disambiguation by short contexts", Comput Hum, 19(3), pp.147~157 (1985)
- Joshi M, "A comparative study of supervised learning as applied to acronym expansion in clinical reports", AMIA Annu Symp Proc, 2006, pp.399~403 (2006)
- Xu H, "Methods for building sense inventories of abbreviations in clinical notes", J Am Med Inform Assoc, 16(1), pp.103~108 (2009)
- Savova GK, "Word sense disambiguation across two domains : biomedical literature and clinical notes", J Biomed Inform, 41(6), pp.1088~1100 (2008)
- Manning CD, "Foundations of statistical natural language processing" (1999), MIT Press
- Aronson AR, "Effective mapping of biomedical text to the UMLS Metathesaurus : the MetaMap program", Proc AMIA Symp, 2001, pp.17~21 (2001)
- McCray AT, "Aggregating UMLS semantic types for reducing conceptual complexity", Stud Health Technol Inform, 84(Pt 1), pp.216~220 (2001)
- Hall M, "The WEKA data mining software : an update", ACM SIGKDD Explor, 11(1), pp.10~18 (2009)
- Meystre SM, "Extracting information from textual documents in the electronic health record : a review of recent research", Yearb Med Inform, 2008, pp.128~144 (2008)
- "NLP research data sets" (2014)[URL]
- "Pittsburgh (PA): Department of Biomedical Informatics" (2014)[URL]
- Liu H, "A study of abbreviations in the UMLS", Proc AMIA Symp, 2001, pp.393~397 (2001)
- Zhou W, "ADAM : another database of abbreviations in MEDLINE", Bioinformatics, 22(22), pp.2813~2818 (2006)
- Melton GB, "Automated identification of synonyms in biomedical acronym sense inventories", Proceedings of the NAACL HLT 2010Second Louhi Workshop on Text and Data Mining of Health Documents, pp.46~52 (2010)
- Resnik P, "Distinguishing systems and distinguishing senses : new evaluation methods for word sense disambiguation", Nat Lang Eng, 5(2), pp.113~133 (1999)
- Leroy G, "Using symbolic knowledge in the UMLS to disambiguate words in small datasets with a naïve Bayes classifier", Stud Health Technol Inform, 107(Pt 1), pp.381~385 (2004)
- Stevenson M, "Disambiguation of biomedical abbreviations", Proceedings of the Workshop on Current Trends in Biomedical Natural Language Processing (BioNLP);, pp.71~79 (2009)
- Liu H, "A multi-aspect comparison study of supervised word sense disambiguation", J Am Med Inform Assoc, 11(4), pp.320~331 (2004)