Distributed Marker Representation for Ambiguous Discourse Markers and Entangled Relations

TL;DR

This paper introduces a novel distributed marker representation (DMR) that captures context-dependent discourse information, improving implicit discourse relation recognition and enhancing interpretability of ambiguous discourse markers.

Contribution

It proposes a data-driven, unsupervised method to learn distributed representations of discourse markers, addressing ambiguity and entanglement in discourse analysis.

Findings

Achieves state-of-the-art performance on implicit discourse relation recognition

Provides better interpretability of discourse marker semantics

Effectively models ambiguity and entanglement among discourse markers

Abstract

Discourse analysis is an important task because it models intrinsic semantic structures between sentences in a document. Discourse markers are natural representations of discourse in our daily language. One challenge is that the markers as well as pre-defined and human-labeled discourse relations can be ambiguous when describing the semantics between sentences. We believe that a better approach is to use a contextual-dependent distribution over the markers to express discourse information. In this work, we propose to learn a Distributed Marker Representation (DMR) by utilizing the (potentially) unlimited discourse marker data with a latent discourse sense, thereby bridging markers with sentence pairs. Such representations can be learned automatically from data without supervision, and in turn provide insights into the data itself. Experiments show the SOTA performance of our DMR on the implicit discourse relation recognition task and strong interpretability. Our method also offers a valuable tool to understand complex ambiguity and entanglement among discourse markers and manually defined discourse relations.