Research On Translation Methods Of Query Items In Chinese-Mongolian Cross-Language Information Retrieval

With the development of the network, the number of internet user in the world has increased rapidly, at the same time, the multilingual state of the Internet resources have become obvious, cross language information retrieval has been a great important research field.There are already many research results in Chinese-English cross-language information retrieval or other languages. However, related research work in the Mongolian information retrieval is still relatively small. Mongolian is one of the most important languages in the world. So the research of Mongolian information retrieval becomes more and more important.As translation in Chinese inquires direct influence on the retrieval performance, we combine with knowledge and statistical method to translate Chinese query item to Mongolian query item. In this paper, we use ICTCLAS of the Chinese sciences academy to identify proper nouns from transliteration loan words, person names, place name in the Chinese query items. We introduce a new phrase segmentation method of phrases segmentation probability as well as a new translation model which includes probability model and sentence length.Experiment results show that introducing the phrase segmentation model, named entity and loanwords recognition has improved significantly translation quality.