基于餐饮数据复杂网络的菜品拉动力分析

发布时间：2018-03-19 05:21

本文选题：菜品拉动力　切入点：复杂网络　出处：《北京交通大学》2017年硕士论文　论文类型：学位论文

【摘要】：随着餐饮业经营逐渐互联网化及餐饮云服务技术的发展,一些拥有相对完善信息管理系统的餐饮商家和服务商积累下大量餐饮交易数据。从餐饮交易数据中挖掘有效信息,对精准营销、销售风险评估等具有重大意义。但现有方法在餐饮历史数据信息挖掘时存在精度低、适用性差等系列问题。为从餐饮交易数据中准确挖掘有效信息,本文致力于以复杂网络理论为基础,提出基于菜品复杂网络的餐饮数据分析方法。在餐饮领域,菜品的拉动力反映了菜品因自身消费增长而拉动其他菜品整体消费增长的能力。本文从餐饮交易数据出发,利用复杂网络相关理论,通过计算和分析菜品拉动力,进行有效信息挖掘。为了实现上述目标,本文首先分析并量化了餐饮交易数据中菜品间作用关系,通过与随机网络模型比较构建了一种符号菜品复杂网络,可以对菜品间的拉动关系和阻碍关系进行定性和定量计算。基于符号菜品网络,本文进一步利用复杂网络传播模型的思想结合餐饮数据特点提出了一种菜品拉动力计算方法。由于餐饮数据中菜品关系的特殊性,现有网络传播模型无法充分反映菜品间关系。本文针对符号网络的特点,通过改变数据判断条件,提出了一种符号网络下的线性阈值模型,并将其应用于菜品拉动力的计算当中。在上述工作的基础上,本文在某餐饮公司提供的真实账单数据上进行了实验,并分别将本文所提方法与餐饮业常用的菜品排序方法以及一种符号网络上的节点排序方法进行了比较和分析,验证了本文方法的合理性和有效性。
[Abstract]:With the development of catering business and the development of catering cloud service technology, some catering merchants and service providers with relatively perfect information management system have accumulated a large amount of catering transaction data. It is of great significance for precision marketing and sales risk assessment. However, the existing methods have some problems such as low precision and poor applicability in the mining of catering historical data. Based on the theory of complex network, this paper puts forward a method of food data analysis based on complex network of dishes. The pulling power of dishes reflects the ability of the whole consumption of other dishes to increase because of their own consumption growth. Based on the catering transaction data, this paper makes use of the theory of complex network to calculate and analyze the pulling power of dishes. In order to achieve the above objectives, this paper first analyzes and quantifies the relationship between the dishes in the catering transaction data, and constructs a complex network of symbolic dishes by comparing with the stochastic network model. Qualitative and quantitative calculations of pull and drag relationships between dishes can be made. Based on the symbolic dishes network, Based on the idea of complex network communication model and the characteristics of catering data, this paper puts forward a method to calculate the pulling power of food. Because of the particularity of the relationship between food and beverage, The existing network propagation model can not fully reflect the relationship between dishes. In this paper, according to the characteristics of symbol network, a linear threshold model based on symbol network is proposed by changing the data judgment condition. On the basis of the above work, this paper has carried on the experiment on the real bill data provided by a catering company. The method proposed in this paper is compared and analyzed with the common food ordering method and a node sorting method on the symbol network. The rationality and effectiveness of the proposed method are verified.
【学位授予单位】：北京交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：O157.5

【参考文献】