uu.seUppsala University Publications
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Exponerade hatkommentarer: En studie av svensk hatkommentarsklassificering
Uppsala University, Disciplinary Domain of Humanities and Social Sciences, Faculty of Languages, Department of Linguistics and Philology.
2016 (Swedish)Independent thesis Basic level (degree of Bachelor), 10 credits / 15 HE creditsStudent thesis
Abstract [sv]

I detta arbete presenteras hatfulla kommentarer på internet som ett sam- hällsproblem som vi bör göra något åt. Webbplatsen Exponerat.net presenteras som en källa till hatfulla kommentarer. Med hjälp av ett förenklande antagande om att de kommentarer som finns på Exponerat kan utgöra en god representation för hatfulla kommentarer på internet konstruerar vi en klassificerare. Klassificeraren utvärderas i två steg; det ena med hjälp av tiofaldig korsvalidering och det andra manuellt. Klassificeraren uppvisar acceptabla precision/recall-värden i det första utvärderingssteget men faller kort i det manuella. Arbetet avslutas med en diskussion om rimligheten i det förenklande antagandet att använda en enda källa. 

Abstract [en]

Hate speech on the internet is a serious issue. This study asks the question: "Is it possible to use machine learning to do something about it?". By using crawled comments from the blog Exponerat.net as a representation of “hate” and comments from the blog Feber.se as “not-hate” we try to construct a classifier. Evaluation in done in two steps; one using 10-fold cross validation and one using manual evaluation methods. The classifier produces an acceptable result in the first step but falls short in the second. The study ends with discussions about if it is even possible to train a classifier using only one source of data. 

Place, publisher, year, edition, pages
2016. , 31 p.
Keyword [sv]
exponerat.net, hatkommentarer, SVM, textklassificering, blogg, näthat
National Category
Language Technology (Computational Linguistics)
Identifiers
URN: urn:nbn:se:uu:diva-296181OAI: oai:DiVA.org:uu-296181DiVA: diva2:936528
Educational program
Bachelor Programme in Language Technology
Supervisors
Examiners
Available from: 2016-06-27 Created: 2016-06-14 Last updated: 2016-06-27Bibliographically approved

Open Access in DiVA

fulltext(272 kB)233 downloads
File information
File name FULLTEXT01.pdfFile size 272 kBChecksum SHA-512
e91ce8c7e1fb0048eb2dd97b2e5d6f7b7406853045429e3e638c062683592af8d5d92963326e4e19fc307be3efafff5e4f285b761092a7c3d52d673815e48fce
Type fulltextMimetype application/pdf

By organisation
Department of Linguistics and Philology
Language Technology (Computational Linguistics)

Search outside of DiVA

GoogleGoogle Scholar
Total: 233 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

urn-nbn

Altmetric score

urn-nbn
Total: 915 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf