基于专利统计分析的用户需求预测方法研究  

Research on user demand forecasting based on patent statistical analysis

作　　者：张凤伟 曹国忠[1,2] 朱玉宁 刘帅 ZHANG Fengwei;CAO Guozhong;ZHU Yuning;LIU shuai(School of Mechanical Engineering,Hebei University of Technology,Tianjin 300130,China;National Technological Innovation Method and Tool Engineering Research Center,Tianjin 300130,China)

机构地区：[1]河北工业大学机械工程学院,天津300130 [2]国家技术创新方法与实施工具工程技术研究中心,天津300130

出　　处：《河北工业大学学报》2020年第5期45-52,共8页Journal of Hebei University of Technology

基　　金：国家自然科学基金(51475137);河北省自然科学基金(E2015202029);河北省高层次人才资助项目(A201500113);河北省高校百名优秀创新人才支持计划(SLRC2017030)。

摘　　要：针对用户需求预测模糊性、抽象性等问题,提出基于专利知识预测用户需求的方法。以Web获取的用户显性需求为关键词检索相关专利,基于特征提取技术利用SAS软件挖掘专利知识单元;通过专利技术单元与用户需求之间的转化模型挖掘用户隐性需求,结合马斯洛需求等级以及需求进化定律预测用户需求;依据Kano属性对用户需求分类,确定用户需求权重,完成用户需求预测,在此基础上构建了基于专利统计分析的用户需求预测模型。最后,以儿童用双轮智能平衡车为例,论证此方法的可行性。Aiming at the problems of ambiguity and abstraction of user demand forecasting,a method for forecast⁃ing user demand based on patent knowledge is proposed.Retrieve related patents by using the explicit user needs obtained from the Web as keywords,and use SAS software to mine patent knowledge units based on feature extraction technology;mine user hidden needs through the transformation model be⁃tween patent technology units and user needs,and combine Maslow′s needs Level and the law of de⁃mand evolution predict user needs;classify user needs according to Kano attributes,determine user de⁃mand weights,and complete user demand predictions.On this basis,a user demand prediction model based on patent statistical analysis is constructed.Finally,take a two-wheeled smart balance bike for children as an example to demonstrate the feasibility of this method.

关 键 词：Web数据 专利知识 数据挖掘 需求获取 需求预测 产品设计 

分 类 号：TB472[一般工业技术—工业设计]