Unstructured Textual Data Analysis

Unstructured [Textual] Data Analysis
Professor Oded Netzer
March 26, 2019
*not a recorded session
Following a highly successful “Intro to Python for Business” inaugural seminar taught by Mattan Griffel on February 27 in the newly launched Big Data for Better Business Seminar Series – with over 100 students and faculty in attendance – we are pleased to announce our second seminar for the spring 2019. Description and links below:
The digitization of information has made a wealth of unstructured data – in the form of text, image audio, and video – available to firms. Such data often include online reviews, customer service call center and chat room conversations, ads, emails, annual reports, and press releases. Additionally, society generates such unstructured data via newspapers articles, movies, songs and books. By some estimates, 80-95% of all business data is unstructured, and most of that unstructured data is text.
These textual data can be used to help businesses and consumers make better decisions. But by itself, all this data is just that: data. For the data to be useful, we have to be able to extract underlying insights in order to measure, track, understand, and interpret the causes and consequences of marketplace behavior.
In this seminar we will discuss how unstructured data, and in particular text data, can be used for business decision making. We will cover case studies ranging from understating competitive market structure from online discussion, to mapping the political affiliation of brands and predicting loan default from loan applications.
Some examples of Professor Netzer’s work on the topic:
https://soundcloud.com/ideasatwork/oded-netzer
https://www8.gsb.columbia.edu/articles/ideas-work/what-makes-idea-creative