Papers with cross-task consistency
Legal Judgment Prediction via Event Extraction with Constraints (2022.acl-long)
Copied to clipboard
| Challenge: | Existing models fail to locate key event information that determines the judgment results. |
| Approach: | They propose an Event-based Prediction Model with constraints that exploits constraints in LJP. |
| Outcome: | The proposed model surpasses existing models on a standard LJP dataset in English and French. |
Constrained Multi-Task Learning for Bridging Resolution (2022.acl-long)
Copied to clipboard
| Challenge: | bridging resolution is the task of recognizing and resolving bridling anaphors in a text. |
| Approach: | They propose a constrained multi-task learning framework for bridging resolution that exploits cross-task consistency constraints to guide the learning process and pre-train the entity coreference model on publicly available coreference data. |
| Outcome: | The proposed model achieves state-of-the-art on three standard evaluation corpora. |
Constrained Multi-Task Learning for Event Coreference Resolution (2021.naacl-main)
Copied to clipboard
| Challenge: | a neural event coreference model is based on a task of determining whether two event mentions refer to the same event . event coreferent tasks require nontrivial tasks such as identifying potential arguments and linking arguments to their event mention. |
| Approach: | They propose a neural event coreference model in which event coreference is jointly trained with five tasks. |
| Outcome: | The proposed model achieves state-of-the-art on the KBP 2017 event coreference dataset. |
SSCR: Iterative Language-Based Image Editing via Self-Supervised Counterfactual Reasoning (2020.emnlp-main)
Copied to clipboard
| Challenge: | Iterative language-based image editing (ILBIE) tasks follow iterative instructions to edit images step by step. data scarcity makes learning the association between vision and language challenging. |
| Approach: | They propose a framework that incorporates counterfactual thinking to overcome data scarcity by combining out-of-distribution instructions with previous images. |
| Outcome: | The proposed model improves the correctness of ILBIE on two IBLIE datasets, even with only 50% of the training data. |