Enhanced optical character recognition technology for Cebu Institute of Technology Library periodical indexing / Larmie T.S. Feliscuzo.
By: Feliscuzo, Larmie T.S [author]
Description: ix, 83 leaves ; 28 cmContent type: text Media type: unmediated Carrier type: volumeSubject(s): Metadata | Libraries -- Information technology | Libraries -- Technological innovations | Digital preservationDDC classification: 025.3 Dissertation note: Dissertation (Doctor of Philosophy in Technology Management) -- Cebu Institute of Technology - University, May 2010. Summary: TITLE: Enhanced Optical Character Recognition Technology for Cebu Institute of Technology Library Periodical Indexing RESEARCHER: Larmie T. Santos ? Feliscuzo DEGREE: Doctor of Philosophy in Technology Management ADVISER: Celbert M. Himang, Ph. D. UNIVERSITY: Cebu Technological University YEAR COMPLETED: May 2010 In this study, an Optical Character Recognition (OCR) Postprocessor that would correct errors encountered in converting scanned images to text of periodical indexes of the Cebu Institute of Technology College Library is developed. The text serves as an input to the Graphical User Interface (GUI)-based periodical indexing software. The postprocessor uses the combination of spellchecking and rule-based approach. It was tested using the metrics for OCR performance namely ? accuracy, precision, insertion, deletion and substitution. Random testing showed that the postprocessor enhanced the precision result of the OCR reader. The software was tested for its acceptability using the revised Theory of Acceptance and Use of Technology (UTAUT). The survey conducted had 171 respondents with the mean opinion score of 4.23 which is equivalent to the criteria Highly Acceptable. It is recommended that continuous study on how to improve OCR performance be made and the GUI-based periodical indexing software be adopted for use.| Item type | Current location | Home library | Call number | Status | Date due | Barcode | Item holds |
|---|---|---|---|---|---|---|---|
THESIS / DISSERTATION
|
GRADUATE LIBRARY | GRADUATE LIBRARY | 025.3 T F3352 2010 (Browse shelf) | Not for loan | CL-T1835 |
Dissertation (Doctor of Philosophy in Technology Management) -- Cebu Institute of Technology - University, May 2010.
TITLE: Enhanced Optical Character Recognition Technology for Cebu Institute of Technology Library Periodical Indexing
RESEARCHER: Larmie T. Santos ? Feliscuzo
DEGREE: Doctor of Philosophy in Technology Management
ADVISER: Celbert M. Himang, Ph. D.
UNIVERSITY: Cebu Technological University
YEAR COMPLETED: May 2010
In this study, an Optical Character Recognition (OCR) Postprocessor that would correct errors encountered in converting scanned images to text of periodical indexes of the Cebu Institute of Technology College Library is developed. The text serves as an input to the Graphical User Interface (GUI)-based periodical indexing software. The postprocessor uses the combination of spellchecking and rule-based approach. It was tested using the metrics for OCR performance namely ? accuracy, precision, insertion, deletion and substitution. Random testing showed that the postprocessor enhanced the precision result of the OCR reader. The software was tested for its acceptability using the revised Theory of Acceptance and Use of Technology (UTAUT). The survey conducted had 171 respondents with the mean opinion score of 4.23 which is equivalent to the criteria Highly Acceptable. It is recommended that continuous study on how to improve OCR performance be made and the GUI-based periodical indexing software be adopted for use.
There are no comments for this item.