基于扩展质量功能展开和网络图的产品大数据分析方法及其应用探讨..docVIP

基于扩展质量功能展幵和网络图的产品 大数据分析方法及其应用探讨 研宄基地 现有大数据分析方法存在侧重算法提升而忽视数据固有关系、难以综合分析网络 形态数据之间连动关系的问题。为解决这些问题，提出了一个基于扩展QFD和网 络图的产品大数据分析方法。该方法由面向数据关系的扩展QFD、基于网络图的 产品大数据关系描述模型和基于该描述模型的产品大数据分析模型组成。该方法 有助于厘清产品各类数据间的固有关系，可将具有复杂结构、多重关系的数据以 清晰的网络结构表现出来，并可综合利用多种大数据分析模型对产品大数据进 行模式探索，从而达到从海量数据中获取关键数据、发现新数据及数据间的新关 系等0标。解决了现奋大数据分析方法忽视数据固奋关系、难以综合分析数据间 连动关系的问题，使数据建模与算法技术更好地结合。 关键词： 大数据分析;质量功能展开;网络图;产品大数据; 唐中君(1969-),男，湖南人，北京工业大学经济与管理学院傅士 生导师;研究方向：需求预测、运营与营销。 收稿曰期:2017-04-13 基金：国家自然科学基金面上项目“基于类比推理的短生命周期无形体验品需求 预测” A Big Data Analytic Method Based on An Extended QFD and Web Graph and Its Application Tang Zhongjun Cui Junfu Yu Haibo School of Economics and Management， Beijing University of Technology, Research Base of Beijing Modern Manufacturing Development; Abstract： Extant big data analytic methods focus on boosting performance of algorithm but ignore to take inherent relationships of data into consideration. And the methods are lack of ability to process web-based data thoroughly. This paper has proposed a big data analytic method based on extended QFD and web graph. The method consists of (1) an data relationship-oriented extended QFD that can identify relationships of big product data categories, (2) a description model of product big data’s categories that is constructed to web form to display the relationships of data categories, and (3) a description model-based big data analytic model which is aimed to recognize patterns in a multi-dimensional way. The proposed method can display the data with complex shape and multiple connections in an explicit way via web form and then explore the big product data by making use of various suitable big data analytic methods for identifying the key data among huge data and finding new data and its relationship. The method can make it possible to combine the algorithm and data modelingmore effectively. Keyword： Big data analytics; Quality function deployment; Web graph; Big product data; Received： 2017-04—13 1引言 数据采集和存储技术的长足进步使企业得以拥有大量与产品有关