Inducing terminologies from text: A case study for the consumer health domain

Abstract

Specialized medical ontologies and terminologies, such as SNOMED CT and the Unified Medical Language System (UMLS), have been successfully leveraged in medical information systems to provide a standard webaccessible medium for interoperability, access, and reuse. However, these clinically oriented terminologies and ontologies cannot provide sufficient support when integrated into consumer-oriented applications, because these applications must "understand" both technical and lay vocabulary. The latter is not part of these specialized terminologies and ontologies. In this article, we propose a two-step approach for building consumer health terminologies from text: 1) automatic extraction of definitions from consumer-oriented articles and web documents, which reflects language in use, rather than relying solely on dictionaries, and 2) learning to map definitions expressed in natural language to terminological knowledge by inducing a syntactic-semantic grammar rather than using hand-written patterns or grammars. We present quantitative and qualitative evaluations of our two-step approach, which show that our framework could be used to induce consumer health terminologies from text.

Query‐performance prediction for effective query routing in domain‐specific repositories

“…In addition to the practical implications of this work, the findings on the characteristics of domain-specific repositories, improvements in query-performance prediction, and application of query-performance prediction for improving retrieval performance in distributed repositories can have implications for several IR-research areas including the design of domain-specific repositories (Bhavnani et al, 2006;Finn & Kushmerick, 2006;Marcial & Hemminger, 2010;Muresan & Klavans, 2013;Pattuelli, 2011;Tang, Yang, & Song, 2013), distributed repositories and federated search (Avrahami et al, 2006;Davis & Lagoze, 2000; Note. We calculated the individual repository specific metrics based on the predictions for each query corpus pair and labeling the predictions as follows: TP = True-Positive: When repository source of a query is correctly identified FP = False-Positive: When a particular repository is selected as the source when it is not TN = True-Negative: When a particular repository is correctly not selected as the source FN = False-Negative: When a repository is incorrectly not selected as the source for a given query Note.…”

“…In addition to the practical implications of this work, the findings on the characteristics of domain-specific repositories, improvements in query-performance prediction, and application of query-performance prediction for improving retrieval performance in distributed repositories can have implications for several IR-research areas including the design of domain-specific repositories (Bhavnani et al, 2006;Finn & Kushmerick, 2006;Marcial & Hemminger, 2010;Muresan & Klavans, 2013;Pattuelli, 2011;Tang, Yang, & Song, 2013), distributed repositories and federated search (Avrahami et al, 2006;Davis & Lagoze, 2000; Paltoglou et al, 2011;Simeoni et al, 2008), query expansion (Alemayehu, 2003;Efthimiadis, 2000;Minker et al, 1973;Shiri & Revie, 2006), query analysis and characteristics (Bashir & Rauber, 2011), and proactive documentrecommendation systems (Liu et al, 2012).…”

A new framework for ethical creation and evaluation of multi-perspective knowledge organization systems

“…Hence, Composers category of the multi-perspective KOS taxonomy includes subcategories such as KOS professionals, domain experts, various communities' representatives, commercial and non-commercial organizations and end users. In addition to eliminating potential biases, such an ethical approach addresses RQ1 and RQ4 above and ensures the system's effectiveness in consumer-oriented applications, as they include lay perspective and vocabulary required for these applications (Muresan and Klavans, 2013).…”

Toward multiviewpoint ontology construction by collaboration of non‐experts and crowdsourcing: The case of the effect of diet on health

“…In a recent review, Simperl and Luczak‐Rösch () assert that today it is generally acknowledged that ontologies should be developed and maintained in a community‐driven manner, with tools providing collaboration platforms, enabling ontology stakeholders to exchange ideas and discuss modeling decisions. Muresan and Klavans () argue that specialized medical terminologies, such as SNOMED CT (http://www.ihtsdo.org/snomed-ct/) cannot provide sufficient support when integrated into consumer‐oriented applications because they do not include the lay vocabulary required for these applications. Hence, numerous methodologies were proposed for collaborative ontology construction by nonexpert users (Kalbasi, Janowicz, Reitsma, Boerboom, & Alesheikh, ; Pereira, ; Simperl & Luczak‐Rösch, ).…”