系统管理学报 ›› 2025, Vol. 34 ›› Issue (1): 40-49.DOI: 10.3969/j.issn.2097-4558.2025.01.004

• 决策科学与运营管理 • 上一篇    下一篇

网络扩散中迭代产品的退市时机与投放种子优化

熊澳,翁克瑞   

  1. 中国地质大学(武汉) 经济管理学院,武汉 430074
  • 收稿日期:2023-03-03 修回日期:2023-09-15 出版日期:2025-01-28 发布日期:2025-01-24
  • 基金资助:
    国家自然科学基金资助项目(72474201,72293572);教育部人文社会科学研究项目(24YJA630101)

Optimization of Seeds Selection and Delisting Time for Successive Generation Products Under Network Diffusion

XIONG Ao,WENG Kerui   

  1. School of Economics and Management, China University of Geosciences, Wuhan 430074, China
  • Received:2023-03-03 Revised:2023-09-15 Online:2025-01-28 Published:2025-01-24

摘要: 科技上的不断创新提高了企业产品推陈出新的速度。在新产品发布时,新产品与旧产品之间存在内部竞争关系,企业需要考虑以下两个问题:一是新产品如何投放,以快速渗透市场;二是旧产品是否即时退市,以为新产品提供更大的市场空间。研究了产品迭代时退市时机与投放种子选择问题:在一个已存在旧产品的社会网络中投放新产品,新旧产品以内部竞争的扩散机制传播其影响力,如何选择旧产品退市时机与新产品投放种子,使得新旧产品的扩散利润最大化。建立了该问题的整数规划模型,刻画了考虑内部竞争的影响力扩散过程,设计了退市时机与种子选择迭代更新的混合贪婪算法(IUHGA)。与经典算法的对比实验显示,IUHGA具有较高的求解质量。研究结果表明:旧产品退市时机与更新产品利润、种子数量、产品最大扩散周期和网络用户聚集度存在一定关系。

关键词: 社会网络, 影响力最大化, 产品迭代, 退市时机

Abstract: In recent years, advancements in science and technology have accelerated the pace of product innovation within enterprises. However, the launch of new products often creates internal competition with existing products in the market. As a result, enterprises must decide how to select new product seeds to quickly penetrate the market and whether old products should be immediately delisted to make space for new products. This paper addresses the problem of selecting new product seeds and determining the optimal delisting time for old product rollover. Specifically, it explores the launch of new products in social networkalready populated with old products. Both new and old products influence each other through a diffusion process driven by internal competition. The goal is to maximize product diffusion profit by strategically selecting new product seeds and timing the delisting of old products. Therefore, it develops an integer programming model, which describes the influence diffusion process considering internal competition, and proposes an iterative updating hybrid greedy algorithm (IUHGA) to find the optimal solution. Comparative experiments with classic algorithms indicate that IUHGA has a higher solution quality. The results show that the timing of delisting old products is closely related to the profit of newer products, the number of selected seeds, the maximum diffusion period, and the aggregation of network users.

Key words: social network, influence maximization, product rollover, delisting time

中图分类号: