COFFEE: A Contrastive Oracle-Free Framework for Event Extraction

TL;DR

This study proposes a new framework, COFFEE, which outperforms state-of-the-art approaches in the oracle-free setting of the event extraction task, as evaluated on two public variants of the ACE05 benchmark.

Abstract

Event extraction is a complex information extraction task that involves extracting events from unstructured text. Prior classification-based methods require comprehensive entity annotations for joint training, while newer generation-based methods rely on heuristic templates containing oracle information such as event type, which is often unavailable in real-world scenarios. In this study, we consider a more realistic setting of this task, namely the Oracle-Free Event Extraction (OFEE) task, where only the input context is given without any oracle information, including event type, event ontology and trigger word. To solve this task, we propose a new framework, called COFFEE, which extracts the events solely based on the document context without referring to any oracle information. In particular, a contrastive selection model is introduced in COFFEE to rectify the generated triggers and handle multi-event instances. The proposed COFFEE outperforms state-of-the-art approaches under the oracle-free setting of the event extraction task, as evaluated on a public event extraction benchmark ACE05.

Figures (5)

• Figure 1: An event extraction example with two events: Transport and Attack. In the 'Transport' event, 'went' is the trigger word, and 'home' is the 'Destination' argument. In the 'Attack' event, 'killed' is the trigger word while 'father-in-law' and 'home' are the 'Agent' and 'Place' arguments, respectively.
• Figure 2: Overview of our proposed COFFEE framework. We train $\mathcal{G}$ to generate trigger candidates $\hat{\mathbf{y}}^t$ that contain trigger word and event type first. These trigger candidates then used to train $\mathcal{S}$ to select the final trigger predictions $\tilde{\mathbf{y}}^t$. In the argument prediction stage, the trained generator is re-used to generate arguments $\tilde{\mathbf{y}}^a$ based on $\tilde{\mathbf{y}}^t$ selected by $\mathcal{S}$. Only the input context $\mathbf{c}$ is required to predict events.
• Figure 3: Example of input and target for the model.
• Figure 4: Effect of threshold in COFFEE framework.
• Figure 5: The influence of the weight $\alpha$ on performance.