Papers by Jill Fain Lehman
Robust Knowledge Graph Completion with Stacked Convolutions and a Student Re-Ranking Network (2021.acl-long)
Copied to clipboard
| Challenge: | Knowledge graphs (KGs) are incomplete because of the large number of benchmark datasets that are not representative of real KGs. |
| Approach: | They develop a deep convolutional network that utilizes textual entity representations to distill the knowledge from the convolution into a student network that re-ranks promising candidate entities. |
| Outcome: | The proposed model outperforms recent methods in a realistic setting where dense connectivity is not guaranteed. |
MedFilter: Improving Extraction of Task-relevant Utterances through Integration of Discourse Structure and Ontological Knowledge (2020.emnlp-main)
Copied to clipboard
| Challenge: | Identifying task-relevant utterances improves performance at downstream medical processing. |
| Approach: | They propose a novel approach that uses task-oriented conversations to improve utterance classification over SOTA models. |
| Outcome: | The proposed model improves on a corpus of 7,000 doctor-patient conversations on 7,000 patient conversations. |
Translational NLP: A New Paradigm and General Principles for Natural Language Processing Research (2021.naacl-main)
Copied to clipboard
| Challenge: | Natural language processing research is often assumed to emerge naturally . many innovations go unapplied and important questions remain unstudied . |
| Approach: | They propose a new paradigm to structure and facilitate the processes by which basic and applied NLP research inform one another. |
| Outcome: | The proposed framework provides a roadmap for developing Translational NLP as a dedicated research area. |
Adapting Event Extractors to Medical Data: Bridging the Covariate Shift (2021.eacl-main)
Copied to clipboard
| Challenge: | a new study examines the performance of event extractors to new domains without labeled data . event extraction is a key sub-task of interest for text understanding pipelines in multiple domains . |
| Approach: | They propose to align marginal distributions of source and target domains to adapt event extractors to new domains . they use clinical notes and doctor-patient conversations as a testbed . |
| Outcome: | The proposed models reach F1 scores of 70.0 and 72.9 on notes and conversations respectively. |