BNNet: Towards A Comprehensive Bangla WordNet

Author:- Saimum Islam, Md Abdul Lotif, Aysha Akther, and Kazi Masudul Alam
Category:- Conference; Year:- 2020
Discipline:- Computer Science & Engineering Discipline
School:- Science, Engineering & Technology School

Abstract

Advanced Natural Language Processing applications are highly dependant on the handiness of linguistic assets such as WordNet. Despite being a popular language in the world, Bangla has an inadequacy of many linguistic resources. Currently available resources are either inadequate for useful operation or not easily accessible for the research community. Constructing a Bangla WordNet again is a significant hurdle as it requires the critical integration of bilingual English-Bangla resources. In this paper, we take the hurdle to combine freely available online materials to build comprehensive English-Bangla and Bangla-English bi-lingual resources. We further build a Bangla WordNet, BNNet, using these resources and map it with the Princeton English WordNet. We again provide useful user interfaces and RESTful API to access and manipulate the Bangla WordNet as well as develop an English-Bangla similarity checker using the proposed system, BNNet.

Read More