Department of Statistics Unitmark
Dietrich College of Humanities and Social Sciences

“He was always very curious,” Prof. Bill Eddy recalled about his former Ph.D.
student, Carnegie Mellon Statistics alumnus Audris Mockus (Ph.D. 94), who is
today the Ericsson-Harlan D. Mills Chair Professor of Digital Archeology and
Evidence Engineering in the Department of Electrical Engineering and Computer
Science at the University of Tennessee, Knoxville. 
“One day, he came in my office
and saw CDs on the shelves.  I told him they contained about fifty
thousand great pictures of planets, which are too many to look at. “But
Audris was intrigued, and we set out to figure out how to deal with so much
data (pictures),” he said. They acquired all of the necessary computer
hardware and built the appropriate software. “We did it for fun, but it resulted
in the 1996 co-authored  ‘An Interactive Icon Index: Images of the Outer
Planets,’ in the Journal of Computational and Graphical Statistics,” Bill said.

 

Audris grew up in Vilnius, Lithuania.   His father, Jonas Mockus, was a
mathematics professor who worked most of his life on optimization problems and,
particularly, the Bayesian approach to global optimization. 
“It led to me thinking of an
academic career as the only sensible choice for independent, creative work,”
Audris said. After graduating from the Moscow Institute of Physics and
Technology with a computer science masters degree, he began looking for a Ph.D.
program.  At the same time, his father asked him to accompany friend and
Carnegie Mellon statistics Prof. Jay Kadane on a flight to a probability conference
in Lithuania. When Audris mentioned he was looking for a Ph.D. program,
Jay suggested he try Carnegie Mellon for Statistics, which he did, and was
accepted. “I enjoyed working with Jay and Bill, and trying to understand
non-traditional data.  The version control data for very large software
systems at AT&T Bell Labs represented a perfect opportunity to understand
how people actually write the source code,” he said.

 

Audris spent 22 years at Bell Labs (now called Nokia), the latter few years in
part-time status upon joining the faculty at the University of Tennessee. 
“I went full-circle to my
original goal of being a professor.  When I started, Bell Labs was
a wonderful place to do research but, by the time I left, it had changed. “I also preferred a university because 

as I get older, I like working with younger people and, also, now I have much more experience and
enjoy sharing it,” he said.

 

Today, he teaches two courses:  the Fundamentals of Digital Archeology and
Evidence Engineering. 
“Lots of data these days are
digital traces generated by software systems people use for work and social
life, and which generate massive amounts of data.  But not the kind of
data statisticians are used to working with.  “The techniques needed to
analyze such data are inspired by methods in archeology; hence digital
archeology.  I go dig up data.  So the course is really about
discovering, retrieving, storing, and analyzing digital data,” he
said. “Evidence engineering starts from my work in software engineering
where software developers write software that satisfies certain
specifications. “That kind of software does not have traditional
functional specifications.  Instead, it has to operate by observing and
reacting to users’ behavior.  In other words, it creates evidence from
these digital traces,” he said.

 

He is presently working on a project to understand how software supply chains and
open source software ecosystems function by collecting and modeling public data
from all open source software.  
His other recent project is
with the university’s Forensic Anthropology Center, renowned nationally as the
“Body Farm,” where he creates tools to annotate and arrange millions of images
for the purpose of modeling the decomposition process. 

 

In December, 2016, he returned to Carnegie Mellon to meet with Computer Science
faculty, and to speak at a CSAFE-mini symposium (forensics related) in the
Statistics Dept. entitled, “Bugs, Bodies, Decomposition and Statistics: Inverse
Problems and Databases.”  He also got to spend time with Bill and
Jay. 
Audris was also planning at that time to return to Lithuania to
see his father, who retired at age 84. “The Social Security clerk could
not believe he could have worked that long.  But he convinced him, and is
enjoying a comfortable retirement,” he said.