Please use this identifier to cite or link to this item: http://hdl.handle.net/1893/27901
Appears in Collections:Computing Science and Mathematics Journal Articles
Peer Review Status: Refereed
Title: Relation Extraction of Medical Concepts Using Categorization and Sentiment Analysis
Author(s): Mondal, Anupam
Cambria, Erik
Das, Dipankar
Hussain, Amir
Bandyopadhyay, Sivaji
Contact Email: ahu@cs.stir.ac.uk
Keywords: Bio-NLP
Category
Medical concept
Medical context
Semantic
Sentiment
Issue Date: 31-Aug-2018
Date Deposited: 4-Oct-2018
Citation: Mondal A, Cambria E, Das D, Hussain A & Bandyopadhyay S (2018) Relation Extraction of Medical Concepts Using Categorization and Sentiment Analysis. Cognitive Computation, 10 (4), pp. 670-685. https://doi.org/10.1007/s12559-018-9567-8
Abstract: In healthcare services, information extraction is the key to understand any corpus-based knowledge. The process becomes laborious when the annotation is done manually for the availability of a large number of text corpora. Hence, future automated extraction systems will be essential for groups of experts such as doctors and medical practitioners as well as non-experts such as patients, to ensure enhanced clinical decision-making for improving healthcare systems. Such extraction systems can be developed using medical concepts and concept-related features as the part of a structured corpus. The latter can assist in assigning the category and sentiment to each of the medical concepts and their lexical contexts. These categories and sentiment assignments constitute semantic relations of medical concepts, with their context, represented by sentences of the corpus. This paper presents a new domain-based knowledge lexicon coupled with a machine learning approach to extract semantic relations. This is done by assigning category and sentiment of the medical concepts and contexts. The categories considered in this research, are diseases, symptoms, drugs, human_anatomy, and miscellaneous medical terms, whereas sentiments are considered as positive and negative. The proposed assignment systems are developed on the top of WordNet of Medical Event (WME) lexicon. The developed lexicon provides medical concepts and their features, namely Parts-Of-Speech (POS), gloss (descriptive explanation), Similar Sentiment Words (SSW), affinity score, gravity score, polarity score, and sentiment. Several well-known supervised classifiers, including Naïve Bayes, Logistic Regression, and support vector-based Sequential Minimal Optimization (SMO) have been applied to evaluate the developed systems. The proposed approaches have resulted in a concepts clustering application by identifying the semantic relations of concepts. The application provides potential exploitation in several domains, such as medical ontologies and recommendation systems.
DOI Link: 10.1007/s12559-018-9567-8
Rights: The publisher does not allow this work to be made publicly available in this Repository. Please use the Request a Copy feature at the foot of the Repository record to request a copy directly from the author. You can only request a copy if you wish to use this work for your own research or private study.
Licence URL(s): http://www.rioxx.net/licenses/under-embargo-all-rights-reserved

Files in This Item:
File Description SizeFormat 
Mondal2018_Article_RelationExtractionOfMedicalCon.pdfFulltext - Published Version1.57 MBAdobe PDFUnder Permanent Embargo    Request a copy

Note: If any of the files in this item are currently embargoed, you can request a copy directly from the author by clicking the padlock icon above. However, this facility is dependent on the depositor still being contactable at their original email address.



This item is protected by original copyright



Items in the Repository are protected by copyright, with all rights reserved, unless otherwise indicated.

The metadata of the records in the Repository are available under the CC0 public domain dedication: No Rights Reserved https://creativecommons.org/publicdomain/zero/1.0/

If you believe that any material held in STORRE infringes copyright, please contact library@stir.ac.uk providing details and we will remove the Work from public display in STORRE and investigate your claim.