ABSTRACT

The Word Sense Disambiguation (WSD) is a process of disambiguating the sense of the text according to its context. Machine translation is one of the challenging task since it requires effective representation of the text to capture semantic relation between Hindi lyrics in English normal language behaviour. This paper focuses on WSD methods to deal with dialects that convert Hindi lyrics to English in its syntactic structure of the words. WSD is a phenomenon for disambiguating the text so that machine would be capable to deduce correct sense of individual given words. WSD is critical for solving natural language tasks such as Machine Translation (MT) and speech processing. The distinguishing proof of significant words in Hindi as the language is not as simple as that of dialects in English. The interpretations of sonnets through the machines are exceptionally essential and deliberate about mind-blowing events. The interpretation of English ballads into other local dialects can turn out to be very straightforward, however, vice-versa is troublesome. This is due to the assortment of structures, classes, and feelings of the local dialects. Various endeavours have been connected far and wide towards the programmed interpretation of ballads from local dialects into English. In this paper, we propose a half breed MT (HBMT) procedure driven by the standard based MT together with measurements based on statistical machine translation (SMT) and rule-based machine translation (RBMT) for WSD in natural script Hindi in English Lyrics. This proposed method improves the semantic and syntactic accuracy of a machine interpretation framework. Finally, the proposed approach result is compared with the machine translation methods such a Google and Microsoft Bing Babylonian and HMT translators provided achieves a better outcome compared to the existing standards.

摘要

词义消歧（WSD）是根据上下文对文本的意义进行消歧的过程。机器翻译是一项具有挑战性的任务，因为它需要有效地表示文本以捕捉英语正常语言行为中印地语歌词之间的语义关系。本文重点介绍 WSD 方法来处理将印地语歌词转换为英语的方言的句法结构。WSD是一种消除文本歧义的现象，以便机器能够推断出单个给定单词的正确含义。WSD 对于解决机器翻译 (MT) 和语音处理等自然语言任务至关重要。印地语中重要词作为语言的区别证明并不像英语中的方言那么简单。通过机器对十四行诗的解释非常重要，并且对令人兴奋的事件进行了深思熟虑。将英语民谣翻译成其他地方方言可能非常简单，但反之则很麻烦。这是由于地方方言的结构、类别和感受的多样性。为了将民谣从地方方言翻译成英语，人们进行了广泛而广泛的努力。在本文中，我们提出了一种由基于标准的 MT 驱动的混血 MT (HBMT) 程序，以及基于统计机器翻译 (SMT) 和基于规则的机器翻译 (RBMT) 的测量结果，用于自然脚本印地语和英语歌词中的 WSD。这种提出的方​​法提高了机器解释框架的语义和句法准确性。最后，