PI: Dejing Dou
We propose to advance biomedical natural language processing by combining ontology-based reasoning, semantic search, and embedding-driven methods to improve information extraction, representation, and consistency detection in biomedical texts.
PI: Dejing Dou
We propose to advance methods for semantic oppositeness detection in natural language, focusing on embedding-based models and contextual representations to improve reasoning and error detection in text.
Dissertations
Conference Papers
Nisansa de Silva and Dejing Dou, "Semantic Oppositeness Embedding Using an Autoencoder-based Learning Model", in Database and Expert Systems Applications, 2019, pp. 159--174. doi: 10.1007/978-3-030-27615-7_12
Nisansa de Silva, Dejing Dou, and Jingshan Huang, "Discovering Inconsistencies in PubMed Abstracts Through Ontology-Based Information Extraction", in Proceedings of the 8th ACM International Conference on Bioinformatics, Computational Biology, and Health Informatics, New York, NY, USA: ACM, 2017, pp. 362--371. doi: 10.1145/3107411.3107452
