面向少按键装置的汉语笔画输入方法  

Chinese Text Input Method based on Strokes for Small Device

作　　者：王忠建[1] 王悦[1] 

机构地区：[1]哈尔滨商业大学计算机与信息工程学院,哈尔滨150028

出　　处：《哈尔滨商业大学学报（自然科学版）》2010年第3期329-333,共5页Journal of Harbin University of Commerce：Natural Sciences Edition

基　　金：黑龙江省自然科学基金资助项目(F2007-12);人事部留学归国人员科技项目择优资助

摘　　要：针对目前少按键设备搭载的汉语输入法多数输入效率较低的现状,在对现在方法进行考察的基础上,提出一种能够自动适应用户输入习惯的笔画输入方法.方法对笔画按键进行了合理的键盘布局并充分利用输入候补的屏幕显示功能,引入分词技术和自动学习机制,使输入法能够通过机器学习自动调节输入词典和联想信息,自动登录新词和删除不用的词汇,使输入法能够随着用户的使用输入效率不断提高.在PC机环境下编程实现了仿真系统,并通过输入实验验证了方法在较小输入词典的状态下,具有较好的输入效率和用户自适应性.Present Chinese input methods usually have a low input efficiency so that a method that can auto-adapt users input habit is proposed by investigating various input methods.The method collocates strokes with keyboard layout reasonable and make use of screen to display more input candidates.The method also uses machine learning and word segmentation technology,so that it can register new words and delete unused words,input efficiency gets an improvement continuously along with the usage.The simulation system is carried out under the PC machine environment.To evaluate performance of the method,input experiments under smaller input dictionary have been done and compared with other input methods,the results indicated that the method have the better input efficiency and user adaptability.

关 键 词：输入方法 机器学习 分词 自适应 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]