| fullrecord |
[{"key": "dc.contributor.author", "value": "Korolainen, Valtteri", "language": null, "element": "contributor", "qualifier": "author", "schema": "dc"}, {"key": "dc.date.accessioned", "value": "2014-08-28T05:58:02Z", "language": "", "element": "date", "qualifier": "accessioned", "schema": "dc"}, {"key": "dc.date.available", "value": "2014-08-28T05:58:02Z", "language": "", "element": "date", "qualifier": "available", "schema": "dc"}, {"key": "dc.date.issued", "value": "2014", "language": null, "element": "date", "qualifier": "issued", "schema": "dc"}, {"key": "dc.identifier.other", "value": "oai:jykdok.linneanet.fi:1444778", "language": null, "element": "identifier", "qualifier": "other", "schema": "dc"}, {"key": "dc.identifier.uri", "value": "https://jyx.jyu.fi/handle/123456789/44127", "language": "", "element": "identifier", "qualifier": "uri", "schema": "dc"}, {"key": "dc.description.abstract", "value": "Erilaiset kieliteknologiasovellukset ovat olleet jo vuosikymmeni\u00e4 arkip\u00e4iv\u00e4ises-s\u00e4 k\u00e4yt\u00f6ss\u00e4. Esimerkiksi ennustava tekstinsy\u00f6tt\u00f6 ja automaattinen korjaus ovat olleet k\u00e4yt\u00f6ss\u00e4 jo vuosikymmeni\u00e4. Puheen tunnistus ja kielen automaattinen k\u00e4\u00e4nt\u00e4minen ovat puolestaan hieman uudempia sovelluksia. Tieteenalana kieli-teknologia on vuosikymmeni\u00e4 vanha, mutta silti koneilla on viel\u00e4 monesti vai-keuksia ymm\u00e4rt\u00e4\u00e4 luonnollisia kieli\u00e4. T\u00e4m\u00e4n tutkimuksen tavoite on kartoittaa koneiden kyky\u00e4 annotoida teksti\u00e4 automaattisesti kun k\u00e4sitelt\u00e4v\u00e4 aineisto sis\u00e4l-t\u00e4\u00e4 slangia. Tutkimus sis\u00e4lt\u00e4\u00e4 empiirisen kokeen automaattisten annotointialgo-rimien toiminnasta. Kielen prosessointi on my\u00f6s nyky\u00e4\u00e4n k\u00e4yt\u00f6ss\u00e4 olevilla al-goritmeilla verrattain raskasta. Osa sovelluksista voidaan kuitenkin suorittaa pilvipalveluissa. Eurooppalaisten kielien prosessointi nykyalgoritmeilla on koh-tuullisen hyv\u00e4ll\u00e4 tasolla verrattuna moniin muihin kieliin. T\u00e4h\u00e4n syyn\u00e4 on huomattavasti laajempi taustaty\u00f6. Vaikka monet sovellukset onnistuisivat usein ymm\u00e4rt\u00e4m\u00e4\u00e4n luonnollista yleiskielt\u00e4, niin slangin prosessointi on huomatta-vasti hankalampaa. P\u00e4\u00e4syyt slangin prosessoinnin haasteellisuudelle ovat slan-gitutkimuksen v\u00e4h\u00e4isyys kieliteknologioihin liittyen sek\u00e4 slangin monesti kompleksisempi luonne. Automaattinen simultaanitulkkaus on jo jossain m\u00e4\u00e4-rin mahdollista nykyaikaisilla kieliteknologiasovelluksilla. Yksi tapa arvioida tietty\u00e4 kieliteknologiaa on analysoida taustalla olevaa sanaluokkaj\u00e4sent\u00e4j\u00e4\u00e4, jonka teht\u00e4v\u00e4 on annotoida tekstifragmentteja. T\u00e4m\u00e4n tutkimuksen tutkimus-ongelmana on selvitt\u00e4\u00e4 n-gram algoritmin suorityskyky muihin k\u00e4yt\u00f6ss\u00e4 ole-viin algoritmeihin n\u00e4hden slangia annotoitaessa. Tilastollisia l\u00e4hestymistapoja k\u00e4ytett\u00e4ess\u00e4 my\u00f6s taustalla oleva manuaalisen j\u00e4sent\u00e4misen laajuus vaikuttaa merkitt\u00e4v\u00e4sti sanaluokkaj\u00e4sent\u00e4j\u00e4n toimintaan. Eurooppalaiset kielet voidaan prosessoida monesti luotettavammin tilastollisilla menetelmill\u00e4, kun taas esi-merkiksi Etel\u00e4-Intian kielet, kuten Hindi, ovat monesti luotettavampia proses-soida s\u00e4\u00e4nt\u00f6ihin perustuvilla menetelmill\u00e4. Englanninkieli voidaan luonnolli-sessa muodossaan annotoida automaattisesti 97% tarkkudella; englanninkieli-sen slangin automaattinen annotointi saavuttaa puolestaan vain 93% tarkkusta-son. Tutkimustuloksista voidaan todeta, ett\u00e4 vaikka algoritmin valinta vaikut-taa osaltaan annotoinnin tarkkuuteen, niin s\u00e4\u00e4nt\u00f6ihin perustuvat menetelm\u00e4t ovat t\u00e4rke\u00e4 lis\u00e4 slangin annotoinnissa. T\u00e4rkein s\u00e4\u00e4nt\u00f6ihin perustuva lis\u00e4mene-telm\u00e4 on sanojen klusterointi.", "language": "fi", "element": "description", "qualifier": "abstract", "schema": "dc"}, {"key": "dc.description.abstract", "value": "Contemporary computers have different capabilities to process natural lan-guages. For example speech recognition and machine translation are both due to study of natural language processing (NLP). Still, machines have some prob-lems of understanding a natural language since words can be ambiguous. Most of the time machines are able to understand the single words. Complete sen-tences are causing more problems. As well, a part of the actual language proc-essing is moved to cloud from local machines due to heavy algorithms that have a high time or space compelexity. English and other European languages have better success rate in NLP solutions than other languages. Mainly this is because of the amount of work and prior analysis done around the language. Even though variety of different NLP solutions exists, they are mainly focused on standard language. Our research contains empirical study which goal is to describe n-gram algorithm suitability in automatic slang annotation context. Slang processing is more problematic than processing standard language, which can be seen in lower accuracy rates. Some of the problems are caused lack of extensive slang analysis when on the other hand some problems are due to complexity of slang. Simultaneous interpreter is one possible solution of up-coming NLP innovations but it has limitations since slang processing is still partly under a development. One way to analyze lingual capabilities of a ma-chine is to evaluate the success rate of Part-of-Speech (POS) tagging. The re-search problem is how n-gram algorithms are performing in slang tagging compared to previously experimented algorithms. As a result of this study it is been found that tagging algorithm selection is in major part of tagger accuracy. In statistical approaches corpus size is remarkably affecting the accuracy as well. Languages are performing differently with different algorithms. For instance, statistical tagging algorithms are mostly having better accuracies in European languages while rule based tagging algorithms are outperforming statistical taggers in South Indian languages. From the POS tagging point of view English slang can be considered as different language from Standard English. While Standard English text can be automatically tagged with success rate of 97% the slang taggers are only fairly reaching 93% success rate. As a conclusion for re-search findings, rule-based approaches are important addition for slang POS taggers. Most important of these kinds of tools is word clustering.", "language": "en", "element": "description", "qualifier": "abstract", "schema": "dc"}, {"key": "dc.description.provenance", "value": "Submitted using Plone Publishing form by Valtteri Korolainen (juvakoro) on 2014-08-28 05:58:01.301258. Form: Pro gradu -lomake (https://kirjasto.jyu.fi/julkaisut/julkaisulomakkeet/pro-gradu-lomake). JyX data: [jyx_publishing-allowed (fi) =True]", "language": "en", "element": "description", "qualifier": "provenance", "schema": "dc"}, {"key": "dc.description.provenance", "value": "Submitted by jyx lomake-julkaisija (jyx-julkaisija@noreply.fi) on 2014-08-28T05:58:02Z\r\nNo. of bitstreams: 2\r\nURN:NBN:fi:jyu-201408282684.pdf: 4926493 bytes, checksum: 6ce3f00373c1e5a96fe6e31594b37fea (MD5)\r\nlicense.html: 4807 bytes, checksum: 337422d4fb3330df8a596cfdc94b1d53 (MD5)", "language": "en", "element": "description", "qualifier": "provenance", "schema": "dc"}, {"key": "dc.description.provenance", "value": "Made available in DSpace on 2014-08-28T05:58:02Z (GMT). No. of bitstreams: 2\r\nURN:NBN:fi:jyu-201408282684.pdf: 4926493 bytes, checksum: 6ce3f00373c1e5a96fe6e31594b37fea (MD5)\r\nlicense.html: 4807 bytes, checksum: 337422d4fb3330df8a596cfdc94b1d53 (MD5)\r\n Previous issue date: 2014", "language": "en", "element": "description", "qualifier": "provenance", "schema": "dc"}, {"key": "dc.format.extent", "value": "1 verkkoaineisto.", "language": null, "element": "format", "qualifier": "extent", "schema": "dc"}, {"key": "dc.format.mimetype", "value": "application/pdf", "language": null, "element": "format", "qualifier": "mimetype", "schema": "dc"}, {"key": "dc.language.iso", "value": "eng", "language": null, "element": "language", "qualifier": "iso", "schema": "dc"}, {"key": "dc.rights", "value": "In Copyright", "language": "en", "element": "rights", "qualifier": null, "schema": "dc"}, {"key": "dc.subject.other", "value": "Part-of-Speech tagging", "language": "", "element": "subject", "qualifier": "other", "schema": "dc"}, {"key": "dc.subject.other", "value": "Hidden-Markov Model", "language": "", "element": "subject", "qualifier": "other", "schema": "dc"}, {"key": "dc.subject.other", "value": "Natural Language Processing", "language": "", "element": "subject", "qualifier": "other", "schema": "dc"}, {"key": "dc.subject.other", "value": "Algorithms", "language": "", "element": "subject", "qualifier": "other", "schema": "dc"}, {"key": "dc.subject.other", "value": "Machine Learning", "language": "", "element": "subject", "qualifier": "other", "schema": "dc"}, {"key": "dc.subject.other", "value": "Language Technologies", "language": "", "element": "subject", "qualifier": "other", "schema": "dc"}, {"key": "dc.title", "value": "Part-of-speech tagging in written slang", "language": null, "element": "title", "qualifier": null, "schema": "dc"}, {"key": "dc.type", "value": "master thesis", "language": null, "element": "type", "qualifier": null, "schema": "dc"}, {"key": "dc.identifier.urn", "value": "URN:NBN:fi:jyu-201408282684", "language": null, "element": "identifier", "qualifier": "urn", "schema": "dc"}, {"key": "dc.type.ontasot", "value": "Pro gradu -tutkielma", "language": "fi", "element": "type", "qualifier": "ontasot", "schema": "dc"}, {"key": "dc.type.ontasot", "value": "Master\u2019s thesis", "language": "en", "element": "type", "qualifier": "ontasot", "schema": "dc"}, {"key": "dc.contributor.faculty", "value": "Informaatioteknologian tiedekunta", "language": "fi", "element": "contributor", "qualifier": "faculty", "schema": "dc"}, {"key": "dc.contributor.faculty", "value": "Faculty of Information Technology", "language": "en", "element": "contributor", "qualifier": "faculty", "schema": "dc"}, {"key": "dc.contributor.department", "value": "Tietojenk\u00e4sittelytieteiden laitos", "language": "fi", "element": "contributor", "qualifier": "department", "schema": "dc"}, {"key": "dc.contributor.department", "value": "Department of Computer Science and Information Systems", "language": "en", "element": "contributor", "qualifier": "department", "schema": "dc"}, {"key": "dc.contributor.organization", "value": "University of Jyv\u00e4skyl\u00e4", "language": "en", "element": "contributor", "qualifier": "organization", "schema": "dc"}, {"key": "dc.contributor.organization", "value": "Jyv\u00e4skyl\u00e4n yliopisto", "language": "fi", "element": "contributor", "qualifier": "organization", "schema": "dc"}, {"key": "dc.subject.discipline", "value": "Tietojenk\u00e4sittelytiede", "language": "fi", "element": "subject", "qualifier": "discipline", "schema": "dc"}, {"key": "dc.date.updated", "value": "2014-08-28T05:58:03Z", "language": "", "element": "date", "qualifier": "updated", "schema": "dc"}, {"key": "dc.type.coar", "value": "http://purl.org/coar/resource_type/c_bdcc", "language": null, "element": "type", "qualifier": "coar", "schema": "dc"}, {"key": "dc.rights.accesslevel", "value": "openAccess", "language": null, "element": "rights", "qualifier": "accesslevel", "schema": "dc"}, {"key": "dc.type.publication", "value": "masterThesis", "language": null, "element": "type", "qualifier": "publication", "schema": "dc"}, {"key": "dc.subject.oppiainekoodi", "value": "601", "language": null, "element": "subject", "qualifier": "oppiainekoodi", "schema": "dc"}, {"key": "dc.subject.yso", "value": "kieliteknologia", "language": null, "element": "subject", "qualifier": "yso", "schema": "dc"}, {"key": "dc.subject.yso", "value": "koneoppiminen", "language": null, "element": "subject", "qualifier": "yso", "schema": "dc"}, {"key": "dc.subject.yso", "value": "algoritmit", "language": null, "element": "subject", "qualifier": "yso", "schema": "dc"}, {"key": "dc.subject.yso", "value": "luonnollinen kieli", "language": null, "element": "subject", "qualifier": "yso", "schema": "dc"}, {"key": "dc.format.content", "value": "fulltext", "language": null, "element": "format", "qualifier": "content", "schema": "dc"}, {"key": "dc.rights.url", "value": "https://rightsstatements.org/page/InC/1.0/", "language": null, "element": "rights", "qualifier": "url", "schema": "dc"}, {"key": "dc.type.okm", "value": "G2", "language": null, "element": "type", "qualifier": "okm", "schema": "dc"}]
|