Computing sciences professor helps reveal hidden patterns with new book on data mining

In writing about how to mine data for trends and patterns, Computing Sciences Professor Megan Squire was able to draw from the lessons and challenges she puts before her students at Elon to help educate a broad range of information technology professionals. 

For her latest book, “Mastering Data Mining with Python: Finding patters hidden in your data,” Squire drew from data sets she uses in her own academic research to teach the steps for analyzing expansive pools of data points through data mining, which she describes to the uninitiated as “using statistics and databases to figure out the patterns you are looking for.”

“One of the reasons this book came together so well is that it mirrors what I do in my classrooms,” Squire said. “The biggest challenge was coming up with enough data sets that weren’t boring, or that haven’t been used a million times before. I was able to use the data from my traditional research in every chapter of my book.”

It’s Squire’s second book for Packt Publishing, a U.K.-based publishing house that focuses on educational and instructional texts for IT professionals. Her first, “Clean Data,” was published in May 2015, with both now available for sale through Amazon.com and other outlets. 

To provide a sense of what data mining is, Squire points to one chapter in the book that looks at creating networks between people based on pieces of data about their own social activity. Data across a 10-year period about a very large pool of software developers can be mined to reveal networks between them that might not be initially apparent, Squire said. 

“We implement a program that can slurp up all that data about a person, and then spit out a network,” Squire said.  

Squire said Packt originally approached her several years ago about writing a data mining book, but at the time, she preferred to focus on what became “Clean Data.” Squire said writing a book is a learning experience for her as well, as she works to explaining lessons and concepts to readers. 

“You learn it so much better when you’re writing it so explicitly,” Squire said. “I learn a lot from writing these books because you can’t gloss over things.”

Looking ahead, Squire said she’d like to explore machine learning in a future book. Machine learning is the process of using statistics and computational methods to predict future outcomes based on a known case. She points to spam filtering as a classic example, but says machine learning has become much more complex and widespread. 

“It’s behind everything cool these days,” Squire said.