English to Malayalam translation
Santhosh Kumar2010, Proceedings of the 1st Amrita ACM-W Celebration on Women in Computing in India - A2CWiC '10
This paper underlines a methodology for translating text from English into the Dravidian language, Malayalam using statistical models. By using a monolingual Malayalam corpus and a bilingual English/Malayalam corpus in the training phase, the machine automatically generates Malayalam translations of English sentences. This paper also discusses a technique to improve the alignment model by incorporating the parts of speech information into the bilingual corpus. Removing the insignificant alignments from the sentence pairs by this approach has ensured better training results. Pre-processing techniques like suffix separation from the Malayalam corpus and stop word elimination from the bilingual corpus also proved to be effective in training. Various handcrafted rules designed for the suffix separation process which can be used as a guideline in implementing suffix separation in Malayalam language are also presented in this paper. The structural difference between the English Malayalam pair is resolved in the decoder by applying the order conversion rules. Experiments conducted on a sample corpus have generated reasonably good Malayalam translations and the results are verified with F measure, BLEU and WER evaluation metrics.
Sign up for access to the world's latest research
Figures (10)
![Table 5. Suffix_keys Out of this classification, words belonging to Swarasandhi and Vyanjanaswara sandhi are of major concern and splitting up such words have more significance in the training process of SMT from English to Malayalam. To implement suffix separation, the category of suffix to be separated has to be identified. In the example ' geavé + sas = rages’, the suffix ‘aes ‘is present in an On setting the word to word alignments in the English Malayalam sentence pair, the inflected Malayalam word is aligned with the English word ‘India’. These alignments add on to the tota alignment weight and in effect reduce the probability rate of the translation of ‘India’ as ‘gnqmj*. For the word ‘India’, the word translation chosen by the decoder is one among the inflected forms and it may not be an apt one that fits the context of the newly translated sentence. To resolve this issue, suffix separation is brought into picture and the corpus with root words is subjected to training. Suffix separation rules are formed by applying sandhi rules in Malyalam in the reverse direction. A classification of sandhi rules based on whether a word ends with a vowel (swaram) or a consonant (vyanjanam) is discussed in [9].](https://figures.academia-assets.com/94732512/table_005.jpg)
