Papers by Adrian Ulges
ManyEnt: A Dataset for Few-shot Entity Typing (2020.coling-main)
Copied to clipboard
| Challenge: | ManyEnt is a benchmark for entity typing in few-shot scenarios . given a sentence with two entity mentions, the model is to assign the entity pair to a relation. |
| Approach: | They propose a benchmark for entity typing models in few-shot scenarios . they propose 256 entity types and 53 entity types for a fine-grain variant . |
| Outcome: | The proposed model can be used in few-shot scenarios with a fine-grain typeset and a coarse-grainability one with 53 entity types. |
An End-to-end Model for Entity-level Relation Extraction using Multi-instance Learning (2021.eacl-main)
Copied to clipboard
| Challenge: | Using a multi-task approach, we extract facts from documents at entity level. |
| Approach: | They propose a multi-task approach that builds upon coreference resolution and gathers relevant signals via multi-instance learning with multi-level representations combining global entity and local mention information. |
| Outcome: | The proposed model is on par with task-specific learning, though more efficient due to shared parameters and training steps. |
Addressing Leakage in Self-Supervised Contextualized Code Retrieval (2022.coling-1)
Copied to clipboard
| Challenge: | a recent study addresses the use of contextualized code retrieval to fill gaps in a partial input program. |
| Approach: | They propose a self-supervised approach to contextualized code retrieval . they propose mutual identifier masking, dedentation, and the selection of syntax-aligned targets . |
| Outcome: | The proposed approach improves retrieval substantially and yields state-of-the-art results for code clone and defect detection. |