Naslov
Speech Act Based Classification of Email Messages in Croatian Language

Autori
Franović, Tin ; Šnajder, Jan

Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni

Izvornik
Proceedings of the Eighth Language Technologies Conference / Erjavec, Tomaž ; Žganec Gros, Jerneja - Ljubljana, 2012, 69-72

Skup
Information Society 2012 - Eighth Language Technologies Conference

Mjesto i datum
Ljubljana, Slovenija, 08.10.2012. - 09.10.2012

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
E-mail classification; speech acts; text classification; Croatian language

Sažetak
Speech acts provide an effective way of summarizing the intended purpose of an email message. In this paper we address the task of speech act classification of email messages in Croatian language. We frame the task as a multilabel text classification problem. We perform thorough evaluation using six machine learning algorithms on message-level, paragraph- level, and sentence-level features. Using message-level features, we achieved an overall best F1 score of over 94%.

Izvorni jezik
Engleski

Znanstvena područja
Računarstvo

POVEZANOST RADA