Measures of semantic similarity and relatedness in the biomedical domain

Abstract

Measures of semantic similarity between concepts are widely used in Natural Language Processing. In this article, we show how six existing domain-independent measures can be adapted to the biomedical domain. These measures were originally based on WordNet, an English lexical database of concepts and relations. In this research, we adapt these measures to the SNOMED-CT ontology of medical concepts. The measures include two path-based measures, and three measures that augment path-based measures with information content statistics from corpora. We also derive a context vector measure based on medical corpora that can be used as a measure of semantic relatedness. These six measures are evaluated against a newly created test bed of 30 medical concept pairs scored by three physicians and nine medical coders. We find that the medical coders and physicians differ in their ratings, and that the context vector measure correlates most closely with the physicians, while the path-based measures and one of the information content measures correlates most closely with the medical coders. We conclude that there is a role both for more flexible measures of relatedness based on information derived from corpora, as well as for measures that rely on existing ontological structures.

Corpus domain effects on distributional semantic modeling of medical terms

“…In the current study, the performance on both similarity and relatedness benchmarks plateaued between 10M and 100M tokens, which is consistent with the findings of another previous study by Pedersen et al (2007), in which we found that the performance of another distributional semantics corpus-based approach to computing semantic relatedness plateaued after the training corpus reached 300 000 clinical notes ($ 66M tokens). These findings may be interpreted as providing additional evidence to show that the size of the corpus used for distributional semantic representations of medical terms does not matter beyond a certain point (e.g.…”

“…Automated approaches for representing the semantic content of terms and similarity and relatedness between them have been widely used in a number of Natural Language Processing (NLP) applications in both general English (Budanitsky and Hirst, 2006;Landauer, 2006;Resnik, 1999;Weeds and Weir, 2005) and specialized terminological domains such as bioinformatics (Ferreira et al, 2013;Lord et al, 2003;Mazandu et al, 2016;Wang et al, 2007;Yang et al, 2012) and medicine (Garla and Brandt, 2012;Lee et al, 2008;Liu et al, 2012;Pakhomov et al, 2010;Pedersen et al, 2007;Sajadi, 2014). A subset of these methods, distributional semantics, relies on the co-occurrence information between words obtained from large corpora of text and makes the assumption that words with similar or related meanings tend to occur in similar contexts.…”

“…biomedical articles). One exception to this was a study by Pedersen et al (2007) that used Mayo Clinic clinical notes as a source of training data; however, while that study had a good domain match, the size of the corpus was relatively modest by current standards ($232M tokens).…”

Comparison of Ontology-Based Semantic-Similarity Measures in the Biomedical Text

“…Pedersen et al [1] Proposed semantic similarity and relatedness in the biomedicine domain, by applied a corpus-based context vector approach to measuring thesimilarity between concepts in SNOMED-CT. Their context vector approach is ontology-free but requires training text, for which, they used text data from Mayo Clinic corpus of medical notes.…”

“…Word Net, UMLS/ICD10) to calculate the distance between the How to cite this paper: Althobaiti, A.F.S. (2017) Comparison of Ontology-Based Semantic-Similarity Measures in the Biome-concept nodes in the ontology tree or hierarchy [1]. The second class of techniques uses training corpora and information content (IC) to estimate the semantic similarity and relatedness between two concepts.…”