Incorporating Domain Knowledge into Natural Language Inference on Clinical Texts

Abstract

Making inference on clinical texts is a task which has not been fully studied. With the newly released, expert annotated MedNLI dataset, this task is being boosted. Compared with open domain data, clinical texts present unique linguistic phenomena, e.g., a large number of medical terms and abbreviations, different written forms for the same medical concept, which make inference much harder. Incorporating domain-specific knowledge is a way to eliminate this problem, in this paper, we assemble a new incorporating medical concept definitions module on the classic enhanced sequential inference model (ESIM), which first extracts the most relevant medical concept for each word, if it exists, then encodes the definition of this medical concept with a bidirectional long short-term network (BiLSTM) to obtain domain-specific definition representations, and attends these definition representations over vanilla word embeddings. The empirical evaluations are conducted to demonstrate that our model improves the prediction performance and achieves a high level of accuracy on the MedNLI dataset. Specifically, the knowledge enhanced word representations contribute significantly to entailment class. INDEX TERMS Attention mechanism, clinical text, medical domain knowledge, natural language inference, word representation.

Modelling Simple and Efficient Data Transformation Scheme for Improving Natural Language Processing

“…The challenges associated with word embedding is discussed by Liang et al [15] where prime filtering operation is carried out towards extracting the logical inference. Consideration of the medical dataset along with domain specific study is carried out by Lu et al [16]. The work of Ludwig et al [17] have presented a mechanism to extract specific action from a given textual contents using Bayesian network.…”

WTMED at MEDIQA 2019: A Hybrid Approach to Biomedical Natural Language Inference

“…On top of this sentence pair modeling scheme, previous studies have independently leveraged syntax (Chen et al, 2016), external knowledge Lu et al, 2019), ensemble methods (Ghaeini et al, 2018b), and language model fine-tuning (Alsentzer et al, 2019) to improve the performance of NLI systems. Nonetheless, to our knowledge, there have been no empirical results on the effect of combining these additions simultaneously.…”

“…The explicit integration of entity-level external knowledge has been used to improve many NLP models' performance (Das et al, 2017;Sun et al, 2018). Domain knowledge has also been demonstrated to be useful for in-domain tasks (Romanov and Shivade, 2018;Lu et al, 2019). Therefore, in addition to generic encoders such as MT-DNN and Tree-LSTM, we further enhance the model with domain-specific knowledge through indirectly leveraging labeled biomedical data for other tasks.…”

Relation Extraction Based on Fusion Dependency Parsing from Chinese EMRs

“…Entity relation in EMRs mainly includes the relation between treatment and disease, treatment and symptom, test and disease, test and symptom, and disease and symptom. At present, the machine learning method is widely used in the field of medical texts [1][2][3][4], including the task of relation extraction of English EMRs [5], and most of the feature selections rely on English medical dictionaries and data sets [6] as well as syntactic analysis [7]. However, the relation extraction of Chinese EMRs is still scarce, which is reflected in two aspects: one is the relation between two specific entities and the other is neglecting the unique features of Chinese EMR texts and sentences.…”