Research On The Recognition Of The Functional Words In Contemporary Uyghur

Along with the widespread use of personal computers and rapid development of the Internet, mankind has entered the information era. Whether it's Uyghur or any other language, whatever type of data we are processing, we cannot be separated from technologies like computers. The study of Uyghur language has become one of the hottest topics in the information processing field, and with the development of computational technologies, from the earliest approaches of word processing to current advanced methods, the computer-linguistic studies of Uyghur has gradually turned to the various levels of natural language processing.In this project, the information processing procedures and researches will be carried out on the lexicological level. Modern Uyghur vocabulary can be divided into two major categories, and they are notional words and functional words. Notional words can dependently become part of speech in sentences and can carry both of lexicological and grammatical meanings, while the functional words serve only for grammatical purposes as an attachment to the notional words, and don't have lexicological meanings. They include for types of functional words: postpositions, conjunctions, modal particles and interjections. In Uyghur, the same functional word may play different or special role in different contexts, and by doing so, may produce ambiguous expressions. The existence of ambiguity hugely decreases the accuracy of recognition of the functional words, hence yields the necessity of semantic analysis. Therefore, disambiguation is one of the focal points of the study of Uyghur functional words.Here are the main arguments and focuses of this paper:1. on the purpose of Uyghur functional word recognition study, we conduct an automatic identification procedure on all the words from the textbooks of Uyghur elementary schools and we provide with detailed analysis. In order to meet the information needs, we established a comparatively adequate data corpus for Uyghur functional words.2. From the perspective of traditional linguistics, we will analyze the grammatical roles and semantic structure of the functional words, and build a database for functional word recognition and its disambiguation rules. By means of theories, technologies and methods of computational linguistics, we will design an automatic functional word recognition system that can recognize the functional words from a given text and provide with detailed interpretive information like semantic annotations and grammatical notes etc.3. The recognition system contains four modules, and each of them has query and statistical functionalities which can help to reduce the amount of time spent on the recognition process, and the system also provides a certain technical platform for future study of the Uyghur language, and also widely applicable for natural language recognition and processing projects. Most importantly, the system will fill the blank of this specific field.4. By test running on the textbooks from five different grades, the system found 8539 functional words, showing an average accuracy of 83.50%. It can be inferred from the test results that the system could play an important role in text processing and analysis, and it can also serve as a new foundation for developing language processing and understanding programs for Uyghur.