Improving Event Representation via Simultaneous Weakly Supervised Contrastive Learning and Clustering
This work addresses event representation learning for NLP tasks, offering an incremental improvement through a hybrid method.
The paper tackles the problem of learning event representations from text by proposing SWCC, a framework that uses simultaneous weakly supervised contrastive learning and clustering to leverage co-occurrence information, resulting in improved performance on Hard Similarity and Transitive Sentence Similarity tasks compared to baselines.
Representations of events described in text are important for various tasks. In this work, we present SWCC: a Simultaneous Weakly supervised Contrastive learning and Clustering framework for event representation learning. SWCC learns event representations by making better use of co-occurrence information of events. Specifically, we introduce a weakly supervised contrastive learning method that allows us to consider multiple positives and multiple negatives, and a prototype-based clustering method that avoids semantically related events being pulled apart. For model training, SWCC learns representations by simultaneously performing weakly supervised contrastive learning and prototype-based clustering. Experimental results show that SWCC outperforms other baselines on Hard Similarity and Transitive Sentence Similarity tasks. In addition, a thorough analysis of the prototype-based clustering method demonstrates that the learned prototype vectors are able to implicitly capture various relations between events.