Part of speech tagging with min‐max modular neural networks
Identifieur interne : 001873 ( Main/Exploration ); précédent : 001872; suivant : 001874Part of speech tagging with min‐max modular neural networks
Auteurs : Qing Ma [Japon] ; Bao Iang Lu [Japon] ; Hitoshi Isahara [Japon] ; Michinori Ichikawa [Japon]Source :
- Systems and Computers in Japan [ 0882-1666 ] ; 2002-06-30.
English descriptors
Abstract
A parts of speech (POS) tagging system using neural networks has been developed by Ma and colleagues. This system can tag unlearned data with a much higher accuracy than that of the Hidden Markov Model (HMM), which is the most popular method of POS tagging. It does so by learning a small Thai corpus on the order of 10,000 words that are ambiguous as to their POSs. However, the three‐layer perceptron used in the system has slow convergence and low learning accuracy even on such a small amount of data. It is therefore difficult to improve accuracy by incrementing the epoch of learning or by increasing the amount of learning data. To solve this problem, the tagging system of this paper makes use of the min‐max modular (M3) neural network of Lu and colleagues. This new system learns faster and has a higher learning accuracy compared with the old one, by decomposing large, complicated POS tagging problems into many smaller, easier problems. Learning accuracy can be improved by using the same learning data and larger data sets can be learned, which results in a much higher tagging accuracy. © 2002 Wiley Periodicals, Inc. Syst Comp Jpn, 33(7): 30–39, 2002; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/scj.1139
Url:
DOI: 10.1002/scj.1139
Affiliations:
Links toward previous steps (curation, corpus...)
- to stream Istex, to step Corpus: 001586
- to stream Istex, to step Curation: 001494
- to stream Istex, to step Checkpoint: 000F87
- to stream Main, to step Merge: 001953
- to stream Main, to step Curation: 001873
Le document en format XML
<record><TEI wicri:istexFullTextTei="biblStruct"><teiHeader><fileDesc><titleStmt><title xml:lang="en">Part of speech tagging with min‐max modular neural networks</title>
<author><name sortKey="Ma, Qing" sort="Ma, Qing" uniqKey="Ma Q" first="Qing" last="Ma">Qing Ma</name>
</author>
<author><name sortKey="Lu, Bao Iang" sort="Lu, Bao Iang" uniqKey="Lu B" first="Bao Iang" last="Lu">Bao Iang Lu</name>
</author>
<author><name sortKey="Isahara, Hitoshi" sort="Isahara, Hitoshi" uniqKey="Isahara H" first="Hitoshi" last="Isahara">Hitoshi Isahara</name>
</author>
<author><name sortKey="Ichikawa, Michinori" sort="Ichikawa, Michinori" uniqKey="Ichikawa M" first="Michinori" last="Ichikawa">Michinori Ichikawa</name>
</author>
</titleStmt>
<publicationStmt><idno type="wicri:source">ISTEX</idno>
<idno type="RBID">ISTEX:0EAF6902DDB2D4F961C7FC639C8FEA5DC15E4A30</idno>
<date when="2002" year="2002">2002</date>
<idno type="doi">10.1002/scj.1139</idno>
<idno type="url">https://api.istex.fr/document/0EAF6902DDB2D4F961C7FC639C8FEA5DC15E4A30/fulltext/pdf</idno>
<idno type="wicri:Area/Istex/Corpus">001586</idno>
<idno type="wicri:Area/Istex/Curation">001494</idno>
<idno type="wicri:Area/Istex/Checkpoint">000F87</idno>
<idno type="wicri:doubleKey">0882-1666:2002:Ma Q:part:of:speech</idno>
<idno type="wicri:Area/Main/Merge">001953</idno>
<idno type="wicri:Area/Main/Curation">001873</idno>
<idno type="wicri:Area/Main/Exploration">001873</idno>
</publicationStmt>
<sourceDesc><biblStruct><analytic><title level="a" type="main" xml:lang="en">Part of speech tagging with min‐max modular neural networks</title>
<author><name sortKey="Ma, Qing" sort="Ma, Qing" uniqKey="Ma Q" first="Qing" last="Ma">Qing Ma</name>
<affiliation wicri:level="1"><country xml:lang="fr">Japon</country>
<wicri:regionArea>Keihanna Human Info‐Communication Research Center, Communications Research Laboratory, Kyoto</wicri:regionArea>
<wicri:noRegion>Kyoto</wicri:noRegion>
</affiliation>
</author>
<author><name sortKey="Lu, Bao Iang" sort="Lu, Bao Iang" uniqKey="Lu B" first="Bao Iang" last="Lu">Bao Iang Lu</name>
<affiliation wicri:level="1"><country xml:lang="fr">Japon</country>
<wicri:regionArea>RIKEN Brain Science Institute, Wako</wicri:regionArea>
<wicri:noRegion>Wako</wicri:noRegion>
</affiliation>
</author>
<author><name sortKey="Isahara, Hitoshi" sort="Isahara, Hitoshi" uniqKey="Isahara H" first="Hitoshi" last="Isahara">Hitoshi Isahara</name>
<affiliation wicri:level="1"><country xml:lang="fr">Japon</country>
<wicri:regionArea>Keihanna Human Info‐Communication Research Center, Communications Research Laboratory, Kyoto</wicri:regionArea>
<wicri:noRegion>Kyoto</wicri:noRegion>
</affiliation>
</author>
<author><name sortKey="Ichikawa, Michinori" sort="Ichikawa, Michinori" uniqKey="Ichikawa M" first="Michinori" last="Ichikawa">Michinori Ichikawa</name>
<affiliation wicri:level="1"><country xml:lang="fr">Japon</country>
<wicri:regionArea>RIKEN Brain Science Institute, Wako</wicri:regionArea>
<wicri:noRegion>Wako</wicri:noRegion>
</affiliation>
</author>
</analytic>
<monogr></monogr>
<series><title level="j">Systems and Computers in Japan</title>
<title level="j" type="abbrev">Syst. Comp. Jpn.</title>
<idno type="ISSN">0882-1666</idno>
<idno type="eISSN">1520-684X</idno>
<imprint><publisher>Wiley Subscription Services, Inc., A Wiley Company</publisher>
<pubPlace>New York</pubPlace>
<date type="published" when="2002-06-30">2002-06-30</date>
<biblScope unit="volume">33</biblScope>
<biblScope unit="issue">7</biblScope>
<biblScope unit="page" from="30">30</biblScope>
<biblScope unit="page" to="39">39</biblScope>
</imprint>
<idno type="ISSN">0882-1666</idno>
</series>
<idno type="istex">0EAF6902DDB2D4F961C7FC639C8FEA5DC15E4A30</idno>
<idno type="DOI">10.1002/scj.1139</idno>
<idno type="ArticleID">SCJ1139</idno>
</biblStruct>
</sourceDesc>
<seriesStmt><idno type="ISSN">0882-1666</idno>
</seriesStmt>
</fileDesc>
<profileDesc><textClass><keywords scheme="KwdEn" xml:lang="en"><term>POS tagging</term>
<term>Thai corpus</term>
<term>min‐max neural network</term>
<term>overlearning.</term>
<term>parallel learning</term>
</keywords>
</textClass>
<langUsage><language ident="en">en</language>
</langUsage>
</profileDesc>
</teiHeader>
<front><div type="abstract" xml:lang="en">A parts of speech (POS) tagging system using neural networks has been developed by Ma and colleagues. This system can tag unlearned data with a much higher accuracy than that of the Hidden Markov Model (HMM), which is the most popular method of POS tagging. It does so by learning a small Thai corpus on the order of 10,000 words that are ambiguous as to their POSs. However, the three‐layer perceptron used in the system has slow convergence and low learning accuracy even on such a small amount of data. It is therefore difficult to improve accuracy by incrementing the epoch of learning or by increasing the amount of learning data. To solve this problem, the tagging system of this paper makes use of the min‐max modular (M3) neural network of Lu and colleagues. This new system learns faster and has a higher learning accuracy compared with the old one, by decomposing large, complicated POS tagging problems into many smaller, easier problems. Learning accuracy can be improved by using the same learning data and larger data sets can be learned, which results in a much higher tagging accuracy. © 2002 Wiley Periodicals, Inc. Syst Comp Jpn, 33(7): 30–39, 2002; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/scj.1139</div>
</front>
</TEI>
<affiliations><list><country><li>Japon</li>
</country>
</list>
<tree><country name="Japon"><noRegion><name sortKey="Ma, Qing" sort="Ma, Qing" uniqKey="Ma Q" first="Qing" last="Ma">Qing Ma</name>
</noRegion>
<name sortKey="Ichikawa, Michinori" sort="Ichikawa, Michinori" uniqKey="Ichikawa M" first="Michinori" last="Ichikawa">Michinori Ichikawa</name>
<name sortKey="Isahara, Hitoshi" sort="Isahara, Hitoshi" uniqKey="Isahara H" first="Hitoshi" last="Isahara">Hitoshi Isahara</name>
<name sortKey="Lu, Bao Iang" sort="Lu, Bao Iang" uniqKey="Lu B" first="Bao Iang" last="Lu">Bao Iang Lu</name>
</country>
</tree>
</affiliations>
</record>
Pour manipuler ce document sous Unix (Dilib)
EXPLOR_STEP=$WICRI_ROOT/Ticri/CIDE/explor/OcrV1/Data/Main/Exploration
HfdSelect -h $EXPLOR_STEP/biblio.hfd -nk 001873 | SxmlIndent | more
Ou
HfdSelect -h $EXPLOR_AREA/Data/Main/Exploration/biblio.hfd -nk 001873 | SxmlIndent | more
Pour mettre un lien sur cette page dans le réseau Wicri
{{Explor lien |wiki= Ticri/CIDE |area= OcrV1 |flux= Main |étape= Exploration |type= RBID |clé= ISTEX:0EAF6902DDB2D4F961C7FC639C8FEA5DC15E4A30 |texte= Part of speech tagging with min‐max modular neural networks }}
This area was generated with Dilib version V0.6.32. |