Napredna pretraga

Pregled bibliografske jedinice broj: 998219

Preliminary Multi-lingual Evaluation of a Question Answering System Based on the Node of Knowledge Method


Čandrlić, Sanja; Ašenbrener Katić, Martina; Jakupović, Alen
Preliminary Multi-lingual Evaluation of a Question Answering System Based on the Node of Knowledge Method // Lecture Notes in Networks and Systems / Arai, Kohei ; Bhatia, Rahul (ur.).
San Francisco, SAD: Springer, 2020. str. 998-1009 doi:10.1007/978-3-030-12388-8_69 (predavanje, međunarodna recenzija, cjeloviti rad (in extenso), znanstveni)


Naslov
Preliminary Multi-lingual Evaluation of a Question Answering System Based on the Node of Knowledge Method

Autori
Čandrlić, Sanja ; Ašenbrener Katić, Martina ; Jakupović, Alen

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

Izvornik
Lecture Notes in Networks and Systems / Arai, Kohei ; Bhatia, Rahul - : Springer, 2020, 998-1009

ISBN
978-3-030-12387-1

Skup
Future of Information and Communication Conference FICC 2019: Advances in Information and Communication

Mjesto i datum
San Francisco, SAD, 14-15.03.2019

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
Question answering (QA) system, Node of knowledge (NOK), Language

Sažetak
The Node of Knowledge (NOK) method is a method for knowledge representation. It is used as a basis for development of formalism for textual knowledge representation (FNOK). FNOK was used for development of a Question Answering (QA) System whose research results are presented in this paper. The QA system is based on storing text in relational databases without losing semantics and it is implemented in Oracle. In this paper, we present the results of preliminary evaluation of the QA system for its independency on the language used. The system was tested using the same set of sentences and questions translated in English, German, Italian and Croatian language. Knowledge in the system was based on 26 simple sentences. 44 questions were set, and the system offered 62 answers for each language (some questions had more than one answer). Research results have shown that this QA system can be used for any of the four languages without the need to change functions and algorithms. It is only necessary to set the dictionary data for each language, implement synonyms and recognize different word forms.

Izvorni jezik
Engleski

Znanstvena područja
Informacijske i komunikacijske znanosti



POVEZANOST RADA


Projekt / tema
13.13.1.2.01
17.14.2.2.03

Ustanove
Veleučilište u Rijeci,
Sveučilište u Rijeci - Odjel za informatiku

Časopis indeksira:


  • Scopus


Citati