A Dynamic Programming Algorithm for the Segmentation of Greek Texts

TL;DR

This paper presents a dynamic programming algorithm for segmenting Greek texts, optimizing for word similarity and segment length, demonstrating high accuracy and promising results in text segmentation tasks.

Contribution

It introduces a novel dynamic programming approach that combines similarity and prior length information for effective Greek text segmentation.

Findings

High segmentation accuracy on Greek texts

Effective integration of similarity and length priors

Promising results for future text segmentation applications

Abstract

In this paper we introduce a dynamic programming algorithm to perform linear text segmentation by global minimization of a segmentation cost function which consists of: (a) within-segment word similarity and (b) prior information about segment length. The evaluation of the segmentation accuracy of the algorithm on a text collection consisting of Greek texts showed that the algorithm achieves high segmentation accuracy and appears to be very innovating and promissing.