Essays on Retail Analytics and Material Information Modeling

Open Access
Author:
Tang, Donghuan
Graduate Program:
Industrial Engineering
Degree:
Master of Science
Document Type:
Master Thesis
Date of Defense:
None
Committee Members:
  • Dr Soundar Kumara, Thesis Advisor
  • Dr Christopher Saldana, Thesis Advisor
  • Dr Akhil Kumar, Thesis Advisor
Keywords:
  • Retail Analytics
  • Clustering
  • Association Rules
  • Material Information Modeling
  • Sustainability Metrics Computational Platform
Abstract:
Large amounts of data have been collected in both the retail business and manufacturing industry. In order for the retailers or manufacturing enterprises to achieve business goals such as managing customer relationships or lowering production costs, traditional statistical analysis and business intelligence tools are not sufficient for the analysis of the collected data. Data mining is considered as a promising approach to the efficient processing and analysis of the data. Data mining techniques include algorithms in machining learning, artificial intelligence, and pattern recognition. Considerable research has been done in data mining applications in the retail or manufacturing industry. This thesis reviews literature in both of these areas. In the retail analytics part of the thesis, the analysis of two retail datasets is presented. Hierarchical clustering and Apriori algorithm are used to cluster the customers and find associations rules within each customer group. In the material information modeling part of the thesis, an information system facilitating the analysis of manufacturing data (Material Information Model) is proposed and the implementation of one analysis application of the system is presented. The implemented software (Sustainability Metrics Computational Platform) consists of a local database storing related manufacturing information, functional modules that handle software functions, and a graphical user interface for user interaction with the software. The retail analysis part considers both the transaction data as well as demographics data. Some interesting observations are generated from this data. For instance, among young couples with children living together and most women not employed cereal is the most frequently bought product type. Furthermore, cereal is the most frequently bought product type within most customer groups.