基于用户偏好的促销产品组合优化研究

发布时间：2018-03-29 18:48

本文选题：产品组合　切入点：用户偏好　出处：《东华大学》2017年硕士论文

【摘要】：随着电子商务发展,互联网在线购物已经成为人们生活的重要组成部分,如何在激烈的市场竞争中生存,需要企业制定良好的策略。面对网络平台多样化的产品,传统的促销产品组合策略已经难以适用。本文结合互联网在线销售特点,提出了基于用户偏好的促销产品组合优化模型。模型采用TOC约束理论下的产品组合模型为基础,利用约束理论简化模型,引入用户偏好系数与流量分配模型,并对用户偏好与流量分配模型进行了深入的研究与分析。最后本文通过实例仿真,来验证模型的可靠性与有效性。首先,本文对研究背景、目的以及意义进行了简要的阐述,对论文所涉及的产品组合问题、用户偏好问题以及隶属于用户偏好的评论情感分析和产品推荐进行了梳理与研究。其次,本文基于对理论背景的分析,确定了流量分配模型与用户偏好的度量方法,同时分别建立了基于用户偏好的单次活动与多次活动的产品组合优化模型。在单次活动下的产品组合优化模型中,考虑企业利润最大化目标,建立了库存约束、流量约束等约束。基于单次活动下的产品组合优化模型,再考虑低库存的运作目标以及产品季节等因素,建立了多次活动下的产品组合优化模型。同时,本文通过主成分分析、相关性分析以及多元线性回归等方法建立流量分配模型,提高了模型对于流量利用率。再通过促销系数研究、评论情感分析与产品推荐研究计算用户偏好系数,提高了模型对于用户偏好的感知能力。最后,本文通过粒子群算法与动态规划算法,求解了单次活动与多次活动下的产品组合模型,并通过与实际情况对比,分析优劣,为企业活动策划提供了建议。本文研究的基于用户偏好的促销产品组合优化模型能够使得企业活动的促销策略更加智能,便捷。在模型实证方面,本文的促销产品组合模型都优于企业实际销售,尤其是多次活动下的产品组合优化模型,在销量、利润、库存水平等方面都优于企业实际销售情况。因此,基于用户偏好的促销产品组合优化模型不仅在理论方面成功结合了用户偏好这一特性,而且在实践方面也有良好的应用价值。
[Abstract]:With the development of electronic commerce, online shopping has become an important part of people's life, how to survive in the fierce market competition, companies need to develop a good strategy. In the face of network platform for a wide variety of products, promotional products combination of traditional strategies has been difficult to apply. Combining with the Internet online sales characteristics, put forward the optimization model of promotional products the combination model based on user preferences. The product portfolio model based on the TOC theory, using the theory of constraint model, introducing the user preference coefficient and flow distribution model, and the user preference and flow distribution model of in-depth research and analysis. Finally, through the example simulation, to verify the reliability and validity of the model first, the research background, purpose and significance are briefly described, portfolio problems related to the Sentiment analysis and product recommendation, user preference and user preferences belonging to analyses and study. Secondly, this paper based on the theoretical analysis on the background, the measurement method of flow distribution model and user preferences, and respectively establishing product portfolio optimization model of single and multiple active user preferences based on activities a combinatorial optimization model in product. In a single event under the consideration of the enterprise profit maximization goal, establishes inventory constraint flow constraints. Product portfolio optimization model based on the single event, and then consider the inventory operation objectives and product seasonal factors, established product portfolio optimization model under multiple activities at the same time, through principal component analysis, correlation analysis and multivariate linear regression method to establish the flow distribution model, improve the model for flow through the utilization. Study on the calculation of user preference promotion coefficient, coefficient of sentiment analysis and recommend product reviews, improve the perception of user preferences for the model. Finally, the particle swarm algorithm and dynamic programming algorithm, product combination model of single and multiple activities for activities, and by comparing with the actual situation, analysis of the advantages and disadvantages, provide the recommendations for enterprise planning activities. This paper studies the user preference promotion product portfolio optimization model can make the activities of the enterprise marketing strategy based on more intelligent and convenient. In the model the empirical aspect, the promotion product portfolio model is better than the actual sales of enterprises, especially the product portfolio optimization model, many activities in the sales, profit that inventory levels are better than the actual sales of enterprises. Therefore, the user preference promotion product portfolio optimization model not only in theory based on It combines the characteristics of user preference and has good application value in practice.

【学位授予单位】：东华大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F274

【相似文献】