English-Hindi Transliteration by Applying Finite Rules to Data before Training Using Statistical Machine Translation

Halder, Mitali; Tyagi, Anant Dev

doi:10.1109/icitcs.2013.6717830

Cited by 7 publications

(3 citation statements)

References 1 publication

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Initially we start our work with a parallel corpus of English-Hindi where Hindi data is in UTF-8 notation. After performing experiments in UTF format, [2] we converted the target side data from UTF to wx format using the UTF8_wx converter to achieve transliteration. Then we move to our final experiment of applying finite rules to the data.…”

Section: Related Workmentioning

confidence: 99%

“…As our target data is in wx-notation, we know that, if a word in source data ends with 'a' will definitely ends with 'A' in when transliterated in Hindi (wx-notation). So, our idea is applying one more 'a' to the end of these words, so that the machine will automatically understand when 'aa' encounters it have to transliterate to 'A' in target notation [2].…”

Section: Related Workmentioning

confidence: 99%

“…Sukanya Dutta et al [1] used word level language recognition and spell check method to improve accuracy of translation. Mitali Halder et al [2] applied finite rules during translation using Stanford Phrasal and Moses application. Shruti Mathur et al [3] used hybrid approach i.e.…”

Section: Literature Reviewmentioning

confidence: 99%

See 2 more Smart Citations

Recognition of Street View Images and Languages and Converting It Into User Understandable Format

2016

IJAERD

View full text Add to dashboard Cite

Camera-enabled mobile devices(Android) are commonly used as interaction platforms for linking the users virtual and physical worlds in commercial applications and numerous research, like serving an augmented reality interface. (OPSs) On-premise signs, a popular form of commercial advertising, are widely used in our living life. The OPSs usually exhibit great visual diversity, accompanied with complex environmental terms(conditions) (e.g., foreground and back-ground clutter). The language barrier among tourists is one of the major difficulty when traveling. The tourists can depend on mobile phone for traveling purposes. Android provides a translation platform for this base, performing 2 times faster than other algorithms. We first proposed an OPS data set, in which images of different businesses are collected from Googles Street View. Further, for addressing the problem of real-world OPS learning then recognition, we developed a probabilistic framework based on the distributional clustering, in which we proposed to exploit the distributional information of each visual feature as a reliable selection criterion for building discriminate OPS models. Our approach is linear, simple and can be executed in a parallel fashion, making it scalable and practical for large-scale multimedia applications.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%