Unsupervised Semantic Frame Induction using Triclustering
This addresses the problem of automatically inducing semantic frames from large-scale text data for natural language processing applications, representing an incremental improvement over existing methods.
The paper tackles unsupervised semantic frame induction by framing it as a triclustering problem, achieving state-of-the-art results on a FrameNet-derived dataset and competitive performance on a verb class clustering task.
We use dependency triples automatically extracted from a Web-scale corpus to perform unsupervised semantic frame induction. We cast the frame induction problem as a triclustering problem that is a generalization of clustering for triadic data. Our replicable benchmarks demonstrate that the proposed graph-based approach, Triframes, shows state-of-the art results on this task on a FrameNet-derived dataset and performing on par with competitive methods on a verb class clustering task.