An Impact of Parts Of Speech Analysis on Automatic Classification of OCR texts

摘要

Automatic classification of Optical Character Reader (OCR) texts is important in applications such as institutional repositories and information retrieval. However it is currently impossible for OCR technology to recognize all characters with accuracy of 100%. Furthermore it is not known whether part of speech (POS) analysis contributes to OCR texts representation in a discriminative way. In this paper we experimentally evaluated POS analysis on OCR texts to formulate an informative feature set. Empirical results indicate that the selection of suitable POS improved classification performance of OCR texts.