Greek Text Segmentation Algorithm

Title: Greek Text Segmentation Algorithm

Research Question: How can we develop an efficient algorithm for segmenting Greek texts?

Methodology: The study proposed a dynamic programming algorithm for segmenting Greek texts. This algorithm minimizes a segmentation cost function, which consists of two factors: (a) within-segment word similarity and (b) prior information about segment length.

Results: The evaluation of the segmentation accuracy of the algorithm on a text collection consisting of Greek texts showed that the algorithm achieved high segmentation accuracy.

Implications: The algorithm's high segmentation accuracy suggests that it is innovative and promising for use in information retrieval. It can be applied to either large texts or a stream of independent, concatenated texts, making it a versatile tool for segmenting Greek texts.

Link to Article: https://arxiv.org/abs/0310041v1 Authors: arXiv ID: 0310041v1