Published December 6, 2019 | Version v1
Conference paper Open

Discriminating Between Similar Languages Using Large Web Corpora

Authors/Creators

  • 1. Lexical Computing

Description

This paper presents a method for discriminating similar lan-guages based on wordlists from large web corpora. The main benefits ofthe approach are language independency, a measure of confidence of theclassification and an easy-to-maintain implementation.The method is evaluated on VarDial 2014 workshop data set. The resultaccuracy is comparable to other methods successfully performing at theworkshop.A tool implementing the method in Python can be obtained from web sitehttp://corpus.tools/.

Files

paper12-suchomel.pdf

Files (234.8 kB)

Name Size
md5:0d0b0d9bf4d3df44965a1e2a24d8a855
234.8 kB Preview Download

Additional details

Funding

European Commission
ELEXIS - European Lexicographic Infrastructure 731015