【摘要】 利用互联网技术及时掌握顾客需求的变化,是现代企业进行产品设计的一项重要任务。在Kano模型中,每个需求只能属于一个需求属性,而客户需求是动态的,会随着职位、偏好等的变化而变化。因此,传统的聚类算法已经不能对客户需求做出更为准确的分析。针对如何掌握顾客需求的方向问题,提出基于贪心迭代搜索的双聚类算法的方法进行研究分析。首先,针对某类产品在线评论进行筛选和过滤,挖掘其中潜在的信息,确定顾客需求要素;其次,利用顾客对不同需求要素的满意度评分数值来建立顾客-需求矩阵;最后,分别使用传统聚类算法和双聚类算法对该矩阵进行研究。研究结果表明,双聚类算法的结果更有效一致,更容易发现数据的相关性,可以为相关行业提供有用的顾客需求信息。更多还原

【Abstract】 It is an important task for modern enterprises to make use of Internet technology to grasp the change of customer requirements timely. Each requirement only belongs to a kind of requirement attribute in the Kano model. However,customer requirement is dynamic and will change with changes in position,crowd,preference,etc. Therefore,traditional clustering algorithms can no longer make a more accurate analysis of customer requirements. Aiming at the problem of how to grasp the direction of customer requirements,this paper proposes a method of biclustering algorithm based on greedy iterative search for research and analysis. Firstly,according to the customer’s screening and filtering of a certain type of product online evaluation,we dig out the potential information in it,and determine the customer requirements elements. Secondly,a customer-demand matrix is developed which is based on customer satisfaction scores for different requirement elements. Finally,the traditional clustering algorithm and the biclustering algorithm are used to study the matrix. The results show that the results of the biclustering algorithm are more effective and consistent,and easier to find the correlation of data,which can provide useful reference value for related industries.更多还原