Social Network Analysis

For the final project in BANA-277, Customer and Social Analytics, our team performed a network analysis for data from a multi-category online store. We utilized the igraph package in R to build our network, visualize the connections between items in the store, and generate network statistics such as hub and authority scores.

The original data contained event level customer data, and many hours were dedicated to transforming the data so that we could get it to work with igraph. Along the way, I developed a small R package that will take in data of this form and return an igraph object. Building this R package was a fun, challenging, and rewarding side project that allowed me to explore the developer side of R, something that I have not had much experience with before.

Our analysis of the data showed the effects that coronavirus has had on consumer’s spending habits: the most frequently purchased items were electronic devices, appliances, and construction equipment. We concluded that this was likely driven by people spending more time at home working remotely, cooking, and working on home-improvement projects. In addition to our network analysis, we constructed a logistic regression to determine what factors influence a user’s propensity to make a purchase of a product. Overall, we had a fun time implementing the skills and techniques that we learned in the class to analyze the data.