藏文文本校对技术研究进展  

Research Progress on the Proofreading Technology of Tibetan Text

作　　者：普哇拉毛 尼玛扎西 道吉扎西 王超 Puwa-Lamao;Nyima-Tashi;Dorje-Tashi;WANG Chao(School of Information Science and Technology,Tibet University,Lhasa 850000,China)

机构地区：[1]西藏大学信息科学技术学院,西藏拉萨850000

出　　处：《高原科学研究》2023年第2期99-104,共6页Plateau Science Research

基　　金：科技创新2030新一代人工智能重大项目(2022ZD0116101);西藏大学校级科研培育计划项目(ZDQMJH22-01).

摘　　要：随着信息技术的迅速发展和互联网的广泛应用,越来越多的藏文信息以电子文档形式存储,电子文档的校对技术已成为自然语言处理中的重要任务。目前藏文文本自动校对的方法主要有基于规则方法、统计方法及深度学习的方法。文章较为系统地分析和阐述了藏文文本自动校对方法,重点介绍了基于规则和深度学习的藏文文本校对模型和方法,探讨了当前藏文文本校对技术中存在的问题及未来展望。With the rapid development of information technology and the broad application of the Internet,more and more Tibetan information is stored in the form of electronic documents,and the proofreading technology of the electronic documents has become a task for scholars.The methods of automatic proofreading of Tibetan texts mainly include the rule-based method,the statistical-based method,and the deep learning-based method.In this paper,the method of automatic proofreading of Tibetan text is systematically analyzed and introduced.Moreover,the model and methods of proofreading of Tibetan text based on rules and deep learning are mainly introduced,and the existing problems in the current proofreading technology of Tibetan text and its prospects are discussed.

关 键 词：自然语言处理 藏文文本 文本校对 深度学习 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]