Automatinis užduočių apimties vetinimas naudojant natūralios kalbos apdorojimo įrankius /

This dissertation explores the possibility of natural language processing tools solving the task effort estimation problem as accurately as 80%. Research is made to justify this claim, where a classic perceptron based machine learning architecture is compared against newer, transformer-based architectures. In this research, the task effort estimation problem is modeled on each of the selected architectures to check which of them are the most accurate. Unlike sentiment or semantic analysis, task effort estimation is not a standard problem for natural language processing tools. However, it can be modeled similarly as a multi-class classification problem which takes a sequence of words, task description, as input data.

Grantor dc:publisher

Institutional Repository of Vilnius University

Year dc:date

2020

Author dc:creator

• Motiejauskas, Jonas,

Contributors dc:contributor

• Vasiliauskas, Vidas

Statement dc:rights

• info:eu-repo/semantics/openAccess

Language dc:language

lit

Repository record dc:identifier

https://repository.vu.lt/VU:ELABAETD81706323&prefLang=en_US

OAI identifier oai:identifier

oai:vu.lt:elaba:81706323