A Transformation Approach between the Similar Language Text
Dawa Yidemucao, Aniwar Tohuti, Qing Yu, Maimaiti Roukeyanmu · Advanced materials research · 2013
This paper first investigates the similarity level between the same family and closer languages (such as Altai family languages) and then examines a transformation between their entries and texts. Cosine similarity measure and dynamic programming (DP) algorithm are used to calculate the similarity and transformation between the source and target languages using a multilingual parallel data set and a function word dictionary. Test data set includes 7,854 paralleled sentences of Chinese, Uyghur, Kazakh and Mongolian various writing systems. Experimental results show that the similarity level of the languages from the same language branch is higher than that between different language branches. And a transformation test focused on the Mongolian language branch showed accuracy of 86.7% for NM to TM and 91.1% for NM to TODO.