Identification problems for OCR characters for text retrieval in ancient books: A case study in the Ancient Collections of the Central Library at UNAM

Abstract

This article describes, in general terms, the problems faced for proper text retrieval through optical character recognition (OCR) in ancient books, by taking a sample of works from the fifteenth to the eighteenth centuries that are protected in the Ancient Collections of the Central Library at UNAM, and digitized by the General Directorate of Libraries. It first presents a conceptual theoretical exposition of OCR and its application in text retrieval to continue with the exemplification of the factors that determine the correct or incorrect identification of the graphemes in these books, by means of some tests applied with Adobe Acrobat 8 Professional and, last, it shows some findings obtained as a result of the analysis and interpretation of the data corresponding to the variables in question.