Llamipa: An Incremental Discourse Parser

TL;DR

Llamipa is an incremental discourse parser based on a fine-tuned large language model, leveraging discourse context for improved performance and capable of processing discourse data incrementally for downstream applications.

Contribution

It introduces the first discourse parsing model fine-tuned on SDRT-style annotations that processes data incrementally, enhancing context utilization and downstream task applicability.

Findings

Substantial performance improvements over encoder-only models

Effective incremental processing of discourse data

First LLM-based discourse parser trained on SDRT annotations

Abstract

This paper provides the first discourse parsing experiments with a large language model(LLM) finetuned on corpora annotated in the style of SDRT (Segmented Discourse Representation Theory Asher, 1993; Asher and Lascarides, 2003). The result is a discourse parser, Llamipa (Llama Incremental Parser), that leverages discourse context, leading to substantial performance gains over approaches that use encoder-only models to provide local, context-sensitive representations of discourse units. Furthermore, it can process discourse data incrementally, which is essential for the eventual use of discourse information in downstream tasks.