Introduction of N-gram into a run-length encoding based ASCII art extraction method

研究成果: Conference contribution

抄録

As ASCII arts can be noise for natural language processing, ASCII art extraction methods can be used to remove them from text. A run-length encoding (RLE) based ASCII art extraction method proposed in our papers uses compression ratio by RLE for recognition of ASCII arts as ASCII arts tend to be compressed small by RLE and non-ASCII arts do not. It is because same characters tend to occur successively in ASCII arts but they do not in non-ASCII arts. Small ASCII arts, however, are not compressed as small as large ASCII arts. In this paper, we add the occurrence number of n-gram of ASCII arts in text into the RLE-based method as a new text attribute to cope with small ASCII arts. Our experimental results show that the new attribute improves the F-measure but it adds language-dependency into the RLE-based method though it is desirable that ASCII art extraction methods are language- independent.

元の言語English
ホスト出版物のタイトルLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
出版者Springer Verlag
ページ28-39
ページ数12
9396
ISBN(印刷物)9783319247991
DOI
出版物ステータスPublished - 2015
イベント15th International Conference on Current Trends in Web Engineering, ICWE 2015 Workshops NLPIT, PEWET, SoWEMine - Rotterdam, Netherlands
継続期間: 2015 6 232015 6 26

出版物シリーズ

名前Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
9396
ISSN(印刷物)03029743
ISSN(電子版)16113349

Other

Other15th International Conference on Current Trends in Web Engineering, ICWE 2015 Workshops NLPIT, PEWET, SoWEMine
Netherlands
Rotterdam
期間15/6/2315/6/26

Fingerprint

Processing

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

これを引用

Suzuki, T. (2015). Introduction of N-gram into a run-length encoding based ASCII art extraction method. : Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (巻 9396, pp. 28-39). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); 巻数 9396). Springer Verlag. https://doi.org/10.1007/978-3-319-24800-4_3

Introduction of N-gram into a run-length encoding based ASCII art extraction method. / Suzuki, Tetsuya.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). 巻 9396 Springer Verlag, 2015. p. 28-39 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); 巻 9396).

研究成果: Conference contribution

Suzuki, T 2015, Introduction of N-gram into a run-length encoding based ASCII art extraction method. : Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). 巻. 9396, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 巻. 9396, Springer Verlag, pp. 28-39, 15th International Conference on Current Trends in Web Engineering, ICWE 2015 Workshops NLPIT, PEWET, SoWEMine, Rotterdam, Netherlands, 15/6/23. https://doi.org/10.1007/978-3-319-24800-4_3
Suzuki T. Introduction of N-gram into a run-length encoding based ASCII art extraction method. : Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). 巻 9396. Springer Verlag. 2015. p. 28-39. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-24800-4_3
Suzuki, Tetsuya. / Introduction of N-gram into a run-length encoding based ASCII art extraction method. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). 巻 9396 Springer Verlag, 2015. pp. 28-39 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{7e77c8c5bc244c1fa4a1c34dee440b00,
title = "Introduction of N-gram into a run-length encoding based ASCII art extraction method",
abstract = "As ASCII arts can be noise for natural language processing, ASCII art extraction methods can be used to remove them from text. A run-length encoding (RLE) based ASCII art extraction method proposed in our papers uses compression ratio by RLE for recognition of ASCII arts as ASCII arts tend to be compressed small by RLE and non-ASCII arts do not. It is because same characters tend to occur successively in ASCII arts but they do not in non-ASCII arts. Small ASCII arts, however, are not compressed as small as large ASCII arts. In this paper, we add the occurrence number of n-gram of ASCII arts in text into the RLE-based method as a new text attribute to cope with small ASCII arts. Our experimental results show that the new attribute improves the F-measure but it adds language-dependency into the RLE-based method though it is desirable that ASCII art extraction methods are language- independent.",
keywords = "ASCII art, Natural language processing, Pattern recognition",
author = "Tetsuya Suzuki",
year = "2015",
doi = "10.1007/978-3-319-24800-4_3",
language = "English",
isbn = "9783319247991",
volume = "9396",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "28--39",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

}

TY - GEN

T1 - Introduction of N-gram into a run-length encoding based ASCII art extraction method

AU - Suzuki, Tetsuya

PY - 2015

Y1 - 2015

N2 - As ASCII arts can be noise for natural language processing, ASCII art extraction methods can be used to remove them from text. A run-length encoding (RLE) based ASCII art extraction method proposed in our papers uses compression ratio by RLE for recognition of ASCII arts as ASCII arts tend to be compressed small by RLE and non-ASCII arts do not. It is because same characters tend to occur successively in ASCII arts but they do not in non-ASCII arts. Small ASCII arts, however, are not compressed as small as large ASCII arts. In this paper, we add the occurrence number of n-gram of ASCII arts in text into the RLE-based method as a new text attribute to cope with small ASCII arts. Our experimental results show that the new attribute improves the F-measure but it adds language-dependency into the RLE-based method though it is desirable that ASCII art extraction methods are language- independent.

AB - As ASCII arts can be noise for natural language processing, ASCII art extraction methods can be used to remove them from text. A run-length encoding (RLE) based ASCII art extraction method proposed in our papers uses compression ratio by RLE for recognition of ASCII arts as ASCII arts tend to be compressed small by RLE and non-ASCII arts do not. It is because same characters tend to occur successively in ASCII arts but they do not in non-ASCII arts. Small ASCII arts, however, are not compressed as small as large ASCII arts. In this paper, we add the occurrence number of n-gram of ASCII arts in text into the RLE-based method as a new text attribute to cope with small ASCII arts. Our experimental results show that the new attribute improves the F-measure but it adds language-dependency into the RLE-based method though it is desirable that ASCII art extraction methods are language- independent.

KW - ASCII art

KW - Natural language processing

KW - Pattern recognition

UR - http://www.scopus.com/inward/record.url?scp=84951019227&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84951019227&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-24800-4_3

DO - 10.1007/978-3-319-24800-4_3

M3 - Conference contribution

AN - SCOPUS:84951019227

SN - 9783319247991

VL - 9396

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 28

EP - 39

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

PB - Springer Verlag

ER -