ResearchBib Share Your Research, Maximize Your Social Impacts
Sign for Notice Everyday Sign up >> Login


Journal: Zbornik Veleučilišta u Rijeci - Journal of the Polytechnic of Rijeka (Vol.9, No. 1)

Publication Date:

Authors : ;

Page : 371-384

Keywords : natural language processing; machine translation; computational linguistics; deep learning; artificial intelligence;

Source : Downloadexternal Find it from : Google Scholarexternal


The paper provides an overview of areas related to the processing of natural languages and their interrelationships, starting from a broader domain such as artificial intelligence, through machine learning, computational linguistics, machine translation methods and especially those based on deep learning. The characteristics, applications, phases and main problems of natural language processing from the lexical, syntactic, semantic, speech and pragmatic perspective are described. The phases of natural language recognition and analysis as well as the natural language generation phase are described. Pre-editing and post-editing procedures using controlled natural languages are given as examples of practices that increase the accuracy and quality of automatic translation and text processing in general. Special focus is given to machine translation and machine translation methods. Approaches to machine translation as statistical, rule-based, example-based, hybrid and deep learning-based approach are described and discussed with regard to their advantages and disadvantages including appropriate application in practice. In the end, still unresolved challenges are given as a direction of future research related to natural language processing and the importance of further development of a deep learning-based approach.

Last modified: 2021-05-24 18:04:49