Saved in:
Bibliographic Details
Main Authors: Sediqin, Mohammadreza, Argamon, Shlomo Engelson
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2501.07723
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912187530870784
author Sediqin, Mohammadreza
Argamon, Shlomo Engelson
author_facet Sediqin, Mohammadreza
Argamon, Shlomo Engelson
contents Segmenting text into Elemental Discourse Units (EDUs) is a fundamental task in discourse parsing. We present a new simple method for identifying EDU boundaries, and hence segmenting them, based on lexical and character n-gram features, using random forest classification. We show that the method, despite its simplicity, outperforms other methods both for segmentation and within a state of the art discourse parser. This indicates the importance of such features for identifying basic discourse elements, pointing towards potentially more training-efficient methods for discourse analysis.
format Preprint
id arxiv_https___arxiv_org_abs_2501_07723
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle ESURF: Simple and Effective EDU Segmentation
Sediqin, Mohammadreza
Argamon, Shlomo Engelson
Computation and Language
Machine Learning
Segmenting text into Elemental Discourse Units (EDUs) is a fundamental task in discourse parsing. We present a new simple method for identifying EDU boundaries, and hence segmenting them, based on lexical and character n-gram features, using random forest classification. We show that the method, despite its simplicity, outperforms other methods both for segmentation and within a state of the art discourse parser. This indicates the importance of such features for identifying basic discourse elements, pointing towards potentially more training-efficient methods for discourse analysis.
title ESURF: Simple and Effective EDU Segmentation
topic Computation and Language
Machine Learning
url https://arxiv.org/abs/2501.07723