Marie A Statistical Approach to Build a Machine Translation System for English Assamese Language Pair

N-gram based Machine Translation for English-Assamese: Two Languages with High Syntactical Dissimilarity

“…P(আছছ )=1. Therefore, আছছ will be put next to get the full translated sentence [12]. There is a provision of getting equal probability also.…”

“…-These are relatively simple and -These are having the ability to scale up. It can be efficiently scaled up by simply increasing the value of n [12]. To make it clearer we can say that an n-gram model predicts by looking into −( −1) , … .…”

“…Translation model aims to generate the sentence d of target language from the sentence c of source language by computing the conditional probability . It can be thought that the target sentence is being generated from source sentence word-by-word [12], [4]. The following example can be considered for better understanding:…”

“…Here a partial alignment hypothesis list is maintained and the search is prompted to start with null hypothesis. That means we get the target sentence from an unknown source word sequence [12]. One example is as follows:…”

“…(অসভৰ ৰাজধানী দিছু ৰ | #), where, # is a place holder for an unknown source word sequence. As the search moves forward, one or more words are added to the hypothesis to extend the list of entries [12]. Example is as follows:…”