This paper presents a transliteration system based on pair Hidden Markov Model (pair HMM) training and Weighted Finite State Transducer (WFST) techniques. Parameters used by WFSTs for transliteration generation are learned from a pair HMM. Parameters from pair-HMM training on English-Russian data sets are found to give better transliteration quality than parameters trained for WFSTs for corresp...

Machine Translation for Indian languages is an emerging research area. Transliteration is one such module that we design while designing a translation system. Transliteration means mapping of source language text into the target language. Simple mapping decreases the efficiency of overall translation system. We propose the use of stemming and part-of-speech tagging for transliteration. The effe...

We propose a framework for transliteration which uses (i) a word-origin detection engine (pre-processing) (ii) a CRF based transliteration engine and (iii) a re-ranking model based on lexiconlookup (post-processing). The results obtained for English-Hindi and EnglishKannada transliteration show that the preprocessing and post-processing modules improve the top-1 accuracy by 7.1%.

Transliteration is a key component of machine translation systems and software internationalization. This paper demonstrates that neural sequence-to-sequence models obtain state of the art or close to state of the art results on existing datasets. In an effort to make machine transliteration accessible, we open source a new Arabic to English transliteration dataset and our trained models.

Machine transliteration is defined as automatic phonetic translation of names across languages. In this paper, we propose synchronous adaptor grammar, a novel nonparametric Bayesian learning approach, for machine transliteration. This model provides a general framework without heuristic or restriction to automatically learn syllable equivalents between languages. The proposed model outperforms ...

Transliteration converts words in a source language (e.g., English) into phonetically equivalent words in a target language (e.g., Vietnamese). Transliteration is therefore used to handle out-of-vocabulary (OOV) words adopted from foreign languages in automatic speech recognition and keyword search systems. While statistical transliteration approaches have been widely adopted, they may not alwa...

Parallel Named Entity pairs are important resources in several NLP tasks, such as, CLIR and MT systems. Further, such pairs may also be used for training transliteration systems, if they are transliterations of each other. In this paper, we profile the performance of a mining methodology in mining parallel named entity transliteration pairs in English and an Indian language, Tamil, leveraging l...

The transliteration has attracted interest of several sections of researchers. Several techniques of transliteration have been developed and used – both statistical based approaches and rule based approaches. In the present method, a simple but effective rule based technique is developed for the transliteration between Bengali script and Meetei Mayek script of written Manipuri text. Typically, ...

Transliteration editors are essential for keying-in language scripts into the computer using QWERTY keyboard. Applications of transliteration editors in the context of Universal Digital Library (UDL) include entry of meta-data and dictionaries for many languages both local and International. In this paper we propose a simple approach for building transliteration editors for International langua...

This paper describes QCRI-MES’s submission on the English-Russian dataset to the Eighth Workshop on Statistical Machine Translation. We generate improved word alignment of the training data by incorporating an unsupervised transliteration mining module to GIZA++ and build a phrase-based machine translation system. For tuning, we use a variation of PRO which provides better weights by optimizing...