服务于制造企业创新绩效评价的知识融合模型研究
发布时间：2018-10-19 18:00
【摘要】：自1995年以来，，知识融合作为一个独立的研究领域，得到了长足的发展。虽然在基本理论构建方面尚不完善，但其在应用领域内的突出表现充分体现了知识融合的价值所在。在众多领域内的研究，已经证明了知识融合能够更好的帮助用户解决问题。因此，研究借助知识融合的手段对制造企业的创新绩效进行评价。使企业能够更好的了解当前创新状况，提升自身创新绩效。 本研究采用粗糙集理论和知识融合方法对制造企业创新绩效影响因素进行了研究。通过研究国内外创新绩效影响因素相关文献，并结合数据的可获得性，选取了创新绩效评价指标。参考近年来学者提出的利用粗糙集构建多知识库的初步想法，采用了模糊C均值聚类算法完成了连续属性的离散化，借助遗传算法对制造企业的创新绩效评价指标进行了属性约简，并引入粗糙集理论中的属性依赖度和属性重要度的概念，构建了多组企业创新绩效评价模型。再将评价模型与对应约简所形成的规则库相结合，共同组成了用于知识融合的知识源。然后，采用模糊积分的方法对多组创新绩效评价模型进行了融合，并建立了包括请求处理模块、信息搜集模块、多知识源构建模块、知识融合模块、结果反馈模块等主要功能模块在内的创新绩效最优决策融合模型。完成了服务于制造企业创新绩效评价的知识融合模型的相关研究工作。 最后，选取了159家在中小企业板上市的制造企业作为样本，对创新绩效最优决策融合模型进行了应用研究。研究结果表明经过知识融合的评价结果能够更加准确的判断企业的创新绩效。同时，还发现了企业规模和知识积累对创新绩效的影响最大，其中企业规模对创新绩效具有正向推动作用，而知识积累则与创新绩效呈现负相关性。
[Abstract]:Since 1995, as an independent research field, knowledge fusion has made great progress. Although the construction of basic theory is not perfect, its outstanding performance in the field of application fully embodies the value of knowledge fusion. Research in many fields has proved that knowledge fusion can better help users solve problems. Therefore, the paper evaluates the innovation performance of manufacturing enterprises by means of knowledge fusion. So that enterprises can better understand the current state of innovation, improve their own innovation performance. In this study, rough set theory and knowledge fusion method are used to study the factors affecting innovation performance of manufacturing enterprises. By studying the related literature of influencing factors of innovation performance at home and abroad, and combining with the availability of data, the evaluation index of innovation performance is selected. Referring to the preliminary idea of using rough set to construct multi-knowledge base, the fuzzy C-means clustering algorithm is used to discretize the continuous attributes. With the help of genetic algorithm, the attribute reduction of innovation performance evaluation index of manufacturing enterprises is carried out, and the concepts of attribute dependency degree and attribute importance degree in rough set theory are introduced, and a multi-group innovation performance evaluation model is constructed. Then the evaluation model is combined with the rule base formed by the corresponding reduction to form a knowledge source for knowledge fusion. Then, the fuzzy integral method is used to fuse the multi-group innovation performance evaluation model, which includes the request processing module, the information collection module, the multi-knowledge source construction module, the knowledge fusion module. Results the optimal decision fusion model of innovation performance including the main functional modules such as feedback module. The related research work of knowledge fusion model serving for the evaluation of innovation performance of manufacturing enterprises is completed. Finally, 159 manufacturing enterprises listed on SME board are selected as samples to study the optimal decision fusion model of innovation performance. The results show that the evaluation results of knowledge fusion can judge the innovation performance more accurately. At the same time, it is found that enterprise size and knowledge accumulation have the greatest impact on innovation performance, in which enterprise scale has a positive role in promoting innovation performance, while knowledge accumulation has a negative correlation with innovation performance.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F425;F273.1;F224
本文编号：2281915
[Abstract]:Since 1995, as an independent research field, knowledge fusion has made great progress. Although the construction of basic theory is not perfect, its outstanding performance in the field of application fully embodies the value of knowledge fusion. Research in many fields has proved that knowledge fusion can better help users solve problems. Therefore, the paper evaluates the innovation performance of manufacturing enterprises by means of knowledge fusion. So that enterprises can better understand the current state of innovation, improve their own innovation performance. In this study, rough set theory and knowledge fusion method are used to study the factors affecting innovation performance of manufacturing enterprises. By studying the related literature of influencing factors of innovation performance at home and abroad, and combining with the availability of data, the evaluation index of innovation performance is selected. Referring to the preliminary idea of using rough set to construct multi-knowledge base, the fuzzy C-means clustering algorithm is used to discretize the continuous attributes. With the help of genetic algorithm, the attribute reduction of innovation performance evaluation index of manufacturing enterprises is carried out, and the concepts of attribute dependency degree and attribute importance degree in rough set theory are introduced, and a multi-group innovation performance evaluation model is constructed. Then the evaluation model is combined with the rule base formed by the corresponding reduction to form a knowledge source for knowledge fusion. Then, the fuzzy integral method is used to fuse the multi-group innovation performance evaluation model, which includes the request processing module, the information collection module, the multi-knowledge source construction module, the knowledge fusion module. Results the optimal decision fusion model of innovation performance including the main functional modules such as feedback module. The related research work of knowledge fusion model serving for the evaluation of innovation performance of manufacturing enterprises is completed. Finally, 159 manufacturing enterprises listed on SME board are selected as samples to study the optimal decision fusion model of innovation performance. The results show that the evaluation results of knowledge fusion can judge the innovation performance more accurately. At the same time, it is found that enterprise size and knowledge accumulation have the greatest impact on innovation performance, in which enterprise scale has a positive role in promoting innovation performance, while knowledge accumulation has a negative correlation with innovation performance.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F425;F273.1;F224
【相似文献】
相关期刊论文 前2条
1 刘海峰;陈琦;刘守生;;从数学建模角度思考《概率论与数理统计》的教改途径[J];科技创新导报;2009年06期
2 ;[J];;年期
相关会议论文 前2条
1 熊青春;陈才;;基于知识融合的数控加工工艺信息平台研究与实现[A];全国先进制造技术高层论坛暨第八届制造业自动化与信息化技术研讨会论文集[C];2009年
2 崔磊;陈清才;郭鸿志;王晓龙;;HowNet与维基百科知识融合中的义类属性自动构建方法[A];中国计算机语言学研究前沿进展（2007-2009）[C];2009年
相关博士学位论文 前2条
1 缑锦;知识融合中若干关键技术研究[D];浙江大学;2005年
2 谢能付;基于语义Web技术的知识融合和同步方法研究[D];中国科学院研究生院（计算技术研究所）;2006年
相关硕士学位论文 前2条
1 李秉哲;服务于制造企业创新绩效评价的知识融合模型研究[D];哈尔滨工业大学;2014年
2 夏跃龙;多源异构民族信息资源知识融合算法研究[D];云南师范大学;2014年
本文编号：2281915
本文链接：https://www.wllwen.com/jingjilunwen/gongyejingjilunwen/2281915.html
