Detailed analysis of implementation of aviation NMT system and the effects of aviation post-processing tools on TDIL tourism corpus

Paul, Saptarshi; Purkhyastha, Bipul shyam

doi:10.1016/j.jksuci.2021.01.016

Cited by 1 publication

(1 citation statement)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The native Indian languages present a huge scope for researchers to work on, specially for technical and unexplored domains. Though untouched till 2017, but work has begun on the aviation domain for Indian languages (English-Bengali pair) and we can conclude with the following points: The first known implementation of English-Indian language NMT based MT system for the aviation domain has been carried out and published with satisfactory results, is documented as "Detailed analysis of successful implementation of aviation NMT system and the effects of aviation postprocessing tools on TDIL tourism corpus", Saptarshi Paul, Bipul Syam Purkaystha, Journal of KING SAUD university-computer and information sciences [12].…”

Section: Conclusion and Future Scopementioning

confidence: 86%

Understanding Aviation English: Challenges and Opportunities in NLP Applications for Indian Languages

Paul¹

2021

Air Traffic Management and Control

Self Cite

View full text Add to dashboard Cite

English is a language that is understood, spoken and used by citizens of a diverse array of countries. The speakers include both native and non-native speakers of English. NLP or Natural Language Processing on the other hand is a branch of computer science that deals with one of the most challenging aspect that a machine can process: dealing with Natural Languages. Natural languages which have evolved over centuries are complete, diverse and highly complex and thus are challenging for a computer system to understand and process. MT or Machine Translation is a more specific part of NLP that translates one natural language to another (English being one of the major researched and sought after languages among them). Though research in the field of NLP and MT has come a long way and many efficient translators are available, still Translation and other NLP applications in specialized domains such as aeronautics are still today a challenge for NLP researchers and developers to achieve. NLP applications are often used in education of English Language, and are therefore a continuous process for Non-Native speakers of English. Non-native English speakers take help of various NLP tools such as E-Dictionary, MT applications and others to better understand the English language and thus learn it better and faster. Aviation English poses a challenge to MT systems and understanding it as a whole requires specialized handling as it has own phonetic pronunciations and terminologies and constituent Out-Of-Vocabulary words. Dealing with Aviation English calls for teaming up of experts from Applied Linguistics, NLP and AI. As a result it becomes a cross-research discipline that covers situations that demand real time use of proper language, e.g. ATC communications. This Paper aims to discuss most recent research methodologies that deals with the Aviation English and reviews the problems posed by it. Being a specialized and structured form of English, the problems are faced by both native and non-native speakers of English Language. Discussion is carried out in the relevant and recent advances of methods in dealing with aviation English language challenges from both, the Human (ICAO/DGCA/AAI) as well as NLP angle. Lastly we have a look at how these challenges are linked to scope for development of applied technologies. Research in experiential Aviation English situations deals with both English for Specific Purposes - ESP (Aeronautics in our case) as well as situations in English as a Foreign Language i.e. EFL (English-Indian language pair).

show abstract

Section: Conclusion and Future Scopementioning

confidence: 86%