北京航空航天大学学报(社会科学版) ›› 2016, Vol. 29 ›› Issue (6): 45-53.DOI: 10.13766/j.bhsk.1008-2204.2014.0578

• 经济与管理 • 上一篇    下一篇

基于ICA K-Means的产品口碑演化聚类与营销分析

李红, 潘娜   

  1. 北京航空航天大学 经济管理学院, 北京 100083
  • 收稿日期:2014-11-13 出版日期:2016-11-25 发布日期:2016-08-24
  • 作者简介:李红(1969-),女,黑龙江哈尔滨人,副教授,博士,研究方向为数据挖掘、社会计算.
  • 基金资助:

    国家自然科学基金资助项目(71471009);教育部人文社科学研究规划基金资助项目(11YJA630044)

Clustering and Marketing Analysis for Products Online Word-of-mouth Activity Series Based on ICA K-Means

LI Hong, PAN Na   

  1. School of Economic and Management, Beijing University of Aeronautics and Astronautics, Beijing 100083, China
  • Received:2014-11-13 Online:2016-11-25 Published:2016-08-24

摘要:

对于产品而言,其在线口碑的活跃度是非常具有代表性的一个指标。在线口碑活跃度的高低,直接揭示产品的生命周期演化模式,对于产品生命周期有全面的了解有助于决策者制定营销计划以及战略。但由于产品在线评论的高维性和复杂性,使得其聚类的难度加大。所以,在普通的K均值算法的基础上引入独立成份分析,对异类产品之间或同类产品在线口碑的活跃度之间进行聚类分析,可以大大降低复杂性和提升聚类准确性;同时深入分析提取出的产品生命周期曲线,有效提升在线口碑信息在电子商务营销管理与决策支持中的作用,深化在线口碑活跃度的管理学视角研究。

关键词: 在线评论, 时间序列聚类, K均值, 独立成分分析, 产品生命周期

Abstract:

For product, online word-of-mouse activity is a very typical index, which reveals life cycle evolution model of product. Understanding the product life cycle helps decision makers to make marketing strategy. It is more difficult to do clustering analysis because the product online comments are high-dimensional and complex. Thus, combining K-means algorithm with independent comment analysis and clustering products by this algorithm can improve the accuracy and reduce complexity in no small measure. Furthermore, in-depth analysis on the product life curve can effectively improve the effect of online word-of-mouth information in e-commerce marketing management and decision support, deepening the research on online reputation activity.

Key words: online reputation, time series clustering, K-means, Independent Component Analysis, life cycle

中图分类号: