基于统计结构模式的特种票据字符识别技术

本文提出一种基于统计结构模式的多体印刷票据的字符识别方法。论述特种票据图像上字符的分割技术,以满足后续识别要求。重点介绍基于复杂指数和四边码的字符预分类技术和基于汉字特征点的汉字字符识别方法,同时针对提取特征点时存在的干扰点问题,提出一种基于笔画方向性的判断准则。实验证明,此识别技术处理速度快,不受环境噪声的影响,能够识别车票常用汉字。

基于非统计偏序结构模式的李赛美治疗情志病证经验研究

目的:基于非统计偏序结构模式数据挖掘方法,分析李赛美教授运用经方治疗情志病证的临床经验。方法:收集李赛美教授门诊以经方诊疗与情志症状密切相关的医案64例,基于形式概念分析理论的非统计偏序结构模式的数据挖掘知识发现方法,建立形式背景,生成属性偏序结构图,并从图中进行模式发现,从而分析李赛美教授运用经方治疗情志病证的经验。结果:情志症状主要有易怒、精神紧张,伴随症状主要是难入睡,常用药是柴胡10 g、炒白术30 g、白芍10 g,常用方是小柴胡汤、柴胡桂枝汤、柴胡加龙骨牡蛎汤、逍遥散、桂枝汤。结论:李李赛美认为情志病证多与肝脾肾有关,治疗重在疏肝、健脾、养阴和血,基于非统计偏序结构模式数据挖掘方法,能有效发现其运用经方治疗情志病证的临床经验。

The Design of PBL Network Teaching System based on Multi-Agent

With the utilization of computer technology and network technology in building network teaching system approaching perfection, PBL teaching model is gradually applied to network teaching. On the basis of PBL teaching theory, this paper proposes a network teaching system based on PBL via network teaching model, using C/S structure and Java language to develop the prototype model and adding multi-agent application level to C/S structure in system designing in order to meet the requirements of the utilization of PBL method in network teaching. This new system successfully implements the innovative application of a problem-based network teaching.