Normalising Medical Concepts in Social Media Texts by Learning Semantic Representation

Abstract

Automatically recognising medical concepts mentioned in social media messages (e.g. tweets) enables several applications for enhancing health quality of people in a community, e.g. real-time monitoring of infectious diseases in population. However, the discrepancy between the type of language used in social media and medical ontologies poses a major challenge. Existing studies deal with this challenge by employing techniques, such as lexical term matching and statistical machine translation. In this work, we handle the medical concept normalisation at the semantic level. We investigate the use of neural networks to learn the transition between layman's language used in social media messages and formal medical language used in the descriptions of medical concepts in a standard ontology. We evaluate our approaches using three different datasets, where social media texts are extracted from Twitter messages and blog posts. Our experimental results show that our proposed approaches significantly and consistently outperform existing effective baselines, which achieved state-of-the-art performance on several medical concept normalisation tasks, by up to 44%.

Deep Neural Models for Medical Concept Normalization in User-Generated Texts

“…Huang and Lu (2015) survey the work done in the organization of biomedical NLP (BioNLP) challenge evaluations up to 2014. These tasks are devoted to the normalization of (1) genes from scientific articles (BioCreative I-III in 2005-2011; (2) (Limsopatham and Collier, 2016) 73.39 ----CNN (Limsopatham and Collier, 2016) 81.41 ----RNN (Limsopatham and Collier, 2016) 79.98 ----Attentional Char-CNN (Niu et al, 2018) 84.65 ----Hierarchical Char-CNN (Han et al, 2017) ----87.7 Ensemble (Sarker et al, 2018) - The 2017 SMM4H shared task (Sarker et al, 2018) was the first effort for the evaluation of NLP methods for the normalization of health-related text from social media on publicly released data. Recent advances in neural networks have been utilized for concept normalization: recent studies have employed convolutional neural networks (Limsopatham and Collier, 2016;Niu et al, 2018) and recurrent neural networks (Belousov et al, 2017;Han et al, 2017).…”

“…Recent works go beyond string matching: these works have tried to view the problem of matching a one-or multi-word expression against a knowledge base as a supervised sequence labeling problem. Limsopatham and Collier (2016) utilized convolutional neural networks (CNNs) for phrase normalization in user reviews, while Tutubalina et al (2018), Han et al (2017), andBelousov et al (2017) applied recurrent neural networks (RNNs) to UGTs, achieving similar results. These works were among the first applications of deep learning techniques to medical concept normalization.…”

Question Answering for Suicide Risk Assessment Using Reddit

“…Without a semantic clustering scheme, where semantically similar sentences and questions are paired based on background knowledge, extracting an answer for a question from an entire Reddit post using mere lexical matches leads to inaccurate responses [18]. Our novel approach of dividing a post into sentences provides better results because 1) sentences in a post that don't convey relevant information are removed improving focus in response extraction and 2) better responses are generated for a question if answer is looked for in the most relevant part of a Reddit post.…”

End-to-End System for Bacteria Habitat Extraction

“…This task is commonly known as named entity normalization or entity linking and various approaches ranging from Levenshtein edit distances to recurrent neural networks have been suggested as the plausible solutions (Tiftikci et al, 2016;Limsopatham and Collier, 2016).…”