首页 | 本学科首页   官方微博 | 高级检索  
     

基于贝叶斯网络的品牌风格语言知识建模方法
引用本文:徐江,王海贤,张克俊. 基于贝叶斯网络的品牌风格语言知识建模方法[J]. 中国机械工程, 2013, 24(19): 2611-2616
作者姓名:徐江  王海贤  张克俊
作者单位:1.东南大学,南京,2100892.浙江大学,杭州,310027
基金项目:国家自然科学基金资助项目(51205059, 61075009);教育部重点实验室开放基金资助项目(CDLS-2010-07)National Natural Science Foundation of China(No. 51205059, 61075009)
摘    要:从风格演化现象和规律出发,描述了品牌意象认知与造型语言的映射关联,借助KISS系统与改进设计模式分析法提取了典型识别造型语义,进而构建了基于贝叶斯网络的品牌造型语言网络模型。该网络在品牌意象认知评价实验基础上,结合贝叶斯评分函数和搜索算法开展学习推理,以模块化思想划分为造型语法子网络、意象词汇子网络和映射关系子网络,并根据节点及条件概率分布趋势来解析意象词汇结构及完形组织特性,挖掘和表征隐含造型语言知识规则。最后,以实例分析证明该建模方法对于风格语言知识获取的有效性和适用性。

关 键 词:品牌  贝叶斯网络  风格语言  知识建模  

Research on Modeling Method of Language Knowledge of Brand Style on Bayesian Network
Affiliation:1.Southeast University,Nanjing,210089 2.Zhejiang University,Hangzhou,310027
Abstract:In view of the evolution phenomenon and its rule of  brand style,the research of  mapping relation between brand image and form language was expound profoundly,and the typical form semantic was extracted with the KISS system and the improved design format  analysis methods,a model of form language was established by the modular thought based on Bayesian network.In the model,firstly the cognition experiments on image cognition were conducted,and the learning and the reasoning of the Bayesian network were advanced through the score functions and search algorithms.Then the form syntax subnetworks U,the imager words subnetworks V and the mapping subnetworks W were divided specifically,the Gestalt organization attributes and the image words structures were interpreted.The tacit knowledge rules were further analysed and discovered according to the relationship among the nodes and the distribution trend of the conditional probability.Finally,a case study was completed to illustrate the validity and the applicability of the modeling method.
Keywords:brand  Bayesian network  style language  knowledge modeling  
点击此处可从《中国机械工程》浏览原始摘要信息
点击此处可从《中国机械工程》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号