Date of Award

2015

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Graduate Group

Operations & Information Management

First Advisor

Kartik Hosanagar

Abstract

Consumers are increasingly spending more time and money online. Business

to consumer e-commerce is growing on average of 20 percent each year and

has reached 1.5 trillion dollars globally in 2014. Given the scale and growth

of consumer online purchase and usage data, firms' ability to understand

and utilize this data is becoming an essential competitive strategy.

But, large-scale data analytics in e-commerce is still at its nascent stage and there

is much to be learned in all aspects of e-commerce. Successful analytics on big data often require a combination of both data mining and econometrics: data mining to reduce or structure

(from unstructured data such as text, photo, and video) large-scale data

and econometric analyses to truly understand and assign causality to interesting

patterns. In my dissertation, I study how firms can better utilize big data

analytics and specific applications of machine learning techniques for improved

e-commerce using theory-driven econometrical and experimental studies. I

show that e-commerce managers can now formulate data-driven strategies for

many aspect of business including cross-selling via recommenders on sales

sites to increasing brand awareness and leads via social media content-engineered-marketing.

These results are readily actionable with far-reaching economical consequences.

Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS