Detecting and analyzing similar words are an important component of natural language processing tasks such as information retrieval, document clustering, making dialog systems, word-sense disambiguation, text summarization and machine translation systems etc. Similar words have notable effect on machine translation. Especially, close languages such as Uyghur, Uzbek, Turkmen, Kazakh, Kyrgyz, Tatar, Azeri and Turkish that belong to the Turkic languages family. There are a huge number of similar words, and these similar words affect the quality of translations. In this article, computer-based methods were suggested to detect similar words in Turkic languages and proposed method has been tested to find similar words in Uzbek and Uyghur.
展开▼
