Style Markers Based on Stop-word List

Varování

Publikace nespadá pod Filozofickou fakultu, ale pod Fakultu informatiky. Oficiální stránka publikace je na webu muni.cz.
Autoři

RYGL Jan MEDVEĎ Marek

Rok publikování 2014
Druh Článek ve sborníku
Konference Eighth Workshop on Recent Advances in Slavonic Natural Language Processing
Fakulta / Pracoviště MU

Fakulta informatiky

Citace
www paper
Obor Informatika
Klíčová slova style marker; stop-word list; corpus
Popis The analysis of author’s characteristic writing style and vocabulary has been used to uncover the identity of authors of documents by both manual linguistic approaches and automatic algorithmic methods. The revealing of the gender, name, or age can help to expose pedophiles in social networks, false product reviews on the Internet servers, or machine translations submitted as manually translated texts. These problems are predominantly solved by a combination of stylometry and machine learning techniques. Since the stylometry focuses on the author’s style, word n-grams cannot be used as a style marker. Stop words are not influenced by a topic of documents, therefore they can be used to create style markers. In this paper, we present a guidance on how to implement stop-word extraction and to include stop-words based style markers into a multilingual classification system based on the stylometry.
Související projekty:

Používáte starou verzi internetového prohlížeče. Doporučujeme aktualizovat Váš prohlížeč na nejnovější verzi.