An impact of linguistic features on automated classification of OCR texts

Gudila Paul Moshi, Lazaro S.P. Busagala, Wataru Ohyama, Tetsushi Wakabayashi, Fumitaka Kimura

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Optical Character reader (OCR) systems can be used in digitizing print documents. OCR texts are generated in the process of digitizing print documents. Usually these texts need to be indexed and organized to simplify their access and retrieval. This can be done by the use of automatic classification techniques. However it is currently impossible for OCR technology to recognize all characters with an accuracy of 100%. Furthermore it is not known whether part of speech (POS) analysis contributes to proper OCR texts representation in a discriminative way. Conventionally, the bag-of-words approach is used in OCR text classification. In this paper we experimentally evaluated POS analysis on OCR texts to formulate an informative feature set. Empirical results indicate that the combination of suitably selected POS improved classification performance of OCR texts.

Original languageEnglish
Title of host publicationProceedings of the 9th IAPR International Workshop on Document Analysis Systems, DAS '10
Pages287-292
Number of pages6
DOIs
Publication statusPublished - Aug 2 2010
Event2010 IAPR Workshop on Document Analysis Systems, DAS 2010 - Boston, MA, United States
Duration: Jun 9 2010Jun 11 2010

Publication series

NameACM International Conference Proceeding Series

Other

Other2010 IAPR Workshop on Document Analysis Systems, DAS 2010
CountryUnited States
CityBoston, MA
Period6/9/106/11/10

Fingerprint

Linguistics
Speech analysis

All Science Journal Classification (ASJC) codes

  • Software
  • Human-Computer Interaction
  • Computer Vision and Pattern Recognition
  • Computer Networks and Communications

Cite this

Moshi, G. P., Busagala, L. S. P., Ohyama, W., Wakabayashi, T., & Kimura, F. (2010). An impact of linguistic features on automated classification of OCR texts. In Proceedings of the 9th IAPR International Workshop on Document Analysis Systems, DAS '10 (pp. 287-292). (ACM International Conference Proceeding Series). https://doi.org/10.1145/1815330.1815367

An impact of linguistic features on automated classification of OCR texts. / Moshi, Gudila Paul; Busagala, Lazaro S.P.; Ohyama, Wataru; Wakabayashi, Tetsushi; Kimura, Fumitaka.

Proceedings of the 9th IAPR International Workshop on Document Analysis Systems, DAS '10. 2010. p. 287-292 (ACM International Conference Proceeding Series).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Moshi, GP, Busagala, LSP, Ohyama, W, Wakabayashi, T & Kimura, F 2010, An impact of linguistic features on automated classification of OCR texts. in Proceedings of the 9th IAPR International Workshop on Document Analysis Systems, DAS '10. ACM International Conference Proceeding Series, pp. 287-292, 2010 IAPR Workshop on Document Analysis Systems, DAS 2010, Boston, MA, United States, 6/9/10. https://doi.org/10.1145/1815330.1815367
Moshi GP, Busagala LSP, Ohyama W, Wakabayashi T, Kimura F. An impact of linguistic features on automated classification of OCR texts. In Proceedings of the 9th IAPR International Workshop on Document Analysis Systems, DAS '10. 2010. p. 287-292. (ACM International Conference Proceeding Series). https://doi.org/10.1145/1815330.1815367
Moshi, Gudila Paul ; Busagala, Lazaro S.P. ; Ohyama, Wataru ; Wakabayashi, Tetsushi ; Kimura, Fumitaka. / An impact of linguistic features on automated classification of OCR texts. Proceedings of the 9th IAPR International Workshop on Document Analysis Systems, DAS '10. 2010. pp. 287-292 (ACM International Conference Proceeding Series).
@inproceedings{a4296d3a284c48fcbec8700881dbb28d,
title = "An impact of linguistic features on automated classification of OCR texts",
abstract = "Optical Character reader (OCR) systems can be used in digitizing print documents. OCR texts are generated in the process of digitizing print documents. Usually these texts need to be indexed and organized to simplify their access and retrieval. This can be done by the use of automatic classification techniques. However it is currently impossible for OCR technology to recognize all characters with an accuracy of 100{\%}. Furthermore it is not known whether part of speech (POS) analysis contributes to proper OCR texts representation in a discriminative way. Conventionally, the bag-of-words approach is used in OCR text classification. In this paper we experimentally evaluated POS analysis on OCR texts to formulate an informative feature set. Empirical results indicate that the combination of suitably selected POS improved classification performance of OCR texts.",
author = "Moshi, {Gudila Paul} and Busagala, {Lazaro S.P.} and Wataru Ohyama and Tetsushi Wakabayashi and Fumitaka Kimura",
year = "2010",
month = "8",
day = "2",
doi = "10.1145/1815330.1815367",
language = "English",
isbn = "9781605587738",
series = "ACM International Conference Proceeding Series",
pages = "287--292",
booktitle = "Proceedings of the 9th IAPR International Workshop on Document Analysis Systems, DAS '10",

}

TY - GEN

T1 - An impact of linguistic features on automated classification of OCR texts

AU - Moshi, Gudila Paul

AU - Busagala, Lazaro S.P.

AU - Ohyama, Wataru

AU - Wakabayashi, Tetsushi

AU - Kimura, Fumitaka

PY - 2010/8/2

Y1 - 2010/8/2

N2 - Optical Character reader (OCR) systems can be used in digitizing print documents. OCR texts are generated in the process of digitizing print documents. Usually these texts need to be indexed and organized to simplify their access and retrieval. This can be done by the use of automatic classification techniques. However it is currently impossible for OCR technology to recognize all characters with an accuracy of 100%. Furthermore it is not known whether part of speech (POS) analysis contributes to proper OCR texts representation in a discriminative way. Conventionally, the bag-of-words approach is used in OCR text classification. In this paper we experimentally evaluated POS analysis on OCR texts to formulate an informative feature set. Empirical results indicate that the combination of suitably selected POS improved classification performance of OCR texts.

AB - Optical Character reader (OCR) systems can be used in digitizing print documents. OCR texts are generated in the process of digitizing print documents. Usually these texts need to be indexed and organized to simplify their access and retrieval. This can be done by the use of automatic classification techniques. However it is currently impossible for OCR technology to recognize all characters with an accuracy of 100%. Furthermore it is not known whether part of speech (POS) analysis contributes to proper OCR texts representation in a discriminative way. Conventionally, the bag-of-words approach is used in OCR text classification. In this paper we experimentally evaluated POS analysis on OCR texts to formulate an informative feature set. Empirical results indicate that the combination of suitably selected POS improved classification performance of OCR texts.

UR - http://www.scopus.com/inward/record.url?scp=77954979705&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77954979705&partnerID=8YFLogxK

U2 - 10.1145/1815330.1815367

DO - 10.1145/1815330.1815367

M3 - Conference contribution

AN - SCOPUS:77954979705

SN - 9781605587738

T3 - ACM International Conference Proceeding Series

SP - 287

EP - 292

BT - Proceedings of the 9th IAPR International Workshop on Document Analysis Systems, DAS '10

ER -