Incorporating Dictionaries into a Neural NetworkArchitecture to Extract COVID-19 MedicalConcepts From Social Media

Abstract We investigate the potential benefit of incorporating dictionary information into a neural network architecture for natural language processing. In particular, we make use ofthis architecture to extract several concepts related to COVID-19 from an on-line medicalforum. We use a sample from the forum to manually curate one dictionary for each concept. In addition, we use MetaMap, which is a tool for extracting biomedical concepts, toidentify a small number of semantic concepts. For a supervised concept extraction task onthe forum data, our best model achieved a macro F1 score of 90%. A major difficulty inmedical concept extraction is obtaining labelled data from which to build supervised models. We investigate the utility of our models to transfer to data derived from a differentsource in two ways. First for producing labels via weak learning and second to performconcept extraction. The dataset we use in this case comprises COVID-19 related tweets andwe achieve an F1 score 81% for symptom concept extraction trained on weakly labelleddata. The utility of our dictionaries is compared with a COVID-19 symptom dictionarythat was constructed directly from Twitter. Further experiments that incorporate BERT anda COVID-19 version of BERTweet demonstrate that the dictionaries provide a commensurate result. Our results show that incorporating small domain dictionaries to deep learningmodels can improve concept extraction tasks. Moreover, models built using dictionariesgeneralize well and are transferable to different datasets on a similar task.