In EDS ansehen

Hybrid-Based Machine Translation Systems

Gespeichert in:
Bibliographische Detailangaben
Titel: Hybrid-Based Machine Translation Systems
Autoren: Mouiad Fadeil Alawneh, Tengku Mohd
Quelle: 5TH INTERNATIONAL CONFERENCE ON COMMUNICATION ENGINEERING AND COMPUTER SCIENCE (CIC-COCOS'24). :312-316
Verlagsinformationen: Cihan University-Erbil, 2024.
Publikationsjahr: 2024
Beschreibung: Machine Translation (MT) is described as the method by which computer software is employed to convert text from one natural language to the other. This process includes taking into consideration each language's grammatical framework and applying examples, rules, as well as grammatical principles to adapt the grammatical structure from the source language (SL) to the target language (TL). In this paper, a method for translating well-formed English sentences into coherent Arabic sentences is introduced, utilizing grammar-based as well as example-based translation techniques to address issues related to word order and grammatical agreement. The methodology suggested is both adaptable and capable of being expanded. The primary benefits include: firstly, a hybrid approach merges the strengths of rule-based (RBMT) as well as example-based (EBMT) methodologies. Secondly, it offers the flexibility to adapt to various languages with only slight adjustments. The OAK Parser analyzes incoming English text to identify the part of speech (POS) for each word, serving as an initial step in translation, utilizing the C# programming language. To maintain data integrity, validation rules are implemented in both the database architecture as well as the programming. A key objective for this system is its capability to function independently, including its seamless integration with broader MT systems for English sentences.
Publikationsart: Article
DOI: 10.24086/cocos2024/paper.1517
Dokumentencode: edsair.doi...........286c43e349d29e2bb13b4d1c95d2696c
Datenbank: OpenAIRE
Beschreibung
Abstract:Machine Translation (MT) is described as the method by which computer software is employed to convert text from one natural language to the other. This process includes taking into consideration each language's grammatical framework and applying examples, rules, as well as grammatical principles to adapt the grammatical structure from the source language (SL) to the target language (TL). In this paper, a method for translating well-formed English sentences into coherent Arabic sentences is introduced, utilizing grammar-based as well as example-based translation techniques to address issues related to word order and grammatical agreement. The methodology suggested is both adaptable and capable of being expanded. The primary benefits include: firstly, a hybrid approach merges the strengths of rule-based (RBMT) as well as example-based (EBMT) methodologies. Secondly, it offers the flexibility to adapt to various languages with only slight adjustments. The OAK Parser analyzes incoming English text to identify the part of speech (POS) for each word, serving as an initial step in translation, utilizing the C# programming language. To maintain data integrity, validation rules are implemented in both the database architecture as well as the programming. A key objective for this system is its capability to function independently, including its seamless integration with broader MT systems for English sentences.
DOI:10.24086/cocos2024/paper.1517