Back to Technology  |  More Profiles
Watch faculty video
Close

Loading the player…


Explore More: Read Online Magazine

Eamonn Keogh

Professor of Computer Science and Engineering
Eamonn Keogh
Tracking the Killer Mosquito
Of the 3,528 different kinds of mosquitoes, less than 5% are vectors of human disease. Moreover, mosquitoes are small—1,000 mosquitoes weigh no more than a penny—and difficult to detect. By combining the strengths of data mining with inexpensive sensors, professor Eamonn Keogh's research allows continual reporting of insect counts and their identification, as well as pinpointing locales of active disease agents for targeted suppression.

Areas of Expertise

Areas of Expertise:
  • Machine learning
  • Data mining
  • Information retrieval
  • Computational entomology
College: Press Releases/Articles: Profile:

Select Honors and Distinctions

  • Vodafone Wireless Innovation Project Winner, 1st Prize (2012)
  • University Scholar (2008–11)
  • Bill and Melinda Gates Foundation Grand Challenge Winner (2010)
  • Best Paper Awards (IEEE ICDM 2007; SIGMOD 2001; WUSS 1997; KDD 1997)
  • Regents’ Faculty Fellowship/Faculty Development Award (2005)
  • BCOE Outstanding Teacher Award (2002–03)
  • National Research Service Award (1998)

Research Summary

Identifying patterns and regularities in large datasets, including datasets of historical cultural artifacts and insect databases

Q&A

Q: Describe your research and its applications.
My research area is in data mining, finding patterns and regularities in large datasets. These datasets could contain anything: financial, scientific, social media, medical data, etc. My particular interests are datasets of historical cultural artifacts (rock art, projectile points and medieval manuscripts) and, more recently, insect data. My work on insect data, which is supported by a grant from the Bill & Melinda Gates Foundation and from the Vodafone Wireless Innovation Project, focuses on the problem of classifying flying insects down to species—and sometimes, sex—using inexpensive sensors. Why do we want to do this? Insects are critical to human health. Insects pollinate at least half the food we eat, but at the same time insect-borne diseases (malaria, yellow fever, etc.) kill at least one million people a year. If we could accurately classify insects in the field in real time, we could plan much cheaper and effective intervention campaigns.

Q: What are the goals of your research?
My overarching goal is to produce ultra-cheap sensors (less than $3 each) that could be produced in bulk and left unattended anywhere (agricultural fields, slums, villages, etc.). These sensors should be able to count and classify all insects that come within a meter, and send hourly reports to a central server, where an appropriate person, such as a farmer, public health worker or medical entomologist, could survey the aggregate data for the region and plan suppression programs.

Q: How is your work important to society?
There already exist technologies to control mosquitoes, however, they are expensive. Our systems could tell the responsible party the optimal place to intervene. By killing more mosquitoes, there will be fewer malaria vectors and less malaria infection, which will save lives in the millions.

Q: What is the biggest myth about your research?
People seem to think that computer scientists spend all their time in front of computers! This may be true for some computer scientists, but I spend a lot of time welding together pieces of an apparatus, chasing insects in the local fields, observing entomologists and other interesting things far away from my computers.

Q: Who is the most influential public voice in your field and what are they saying?
Bill and Melinda Gates are almost unique in believing that malaria can be eradicated, not just controlled. Since about 3.3 billion people—half the world’s population—are at risk for malaria, eradicating the disease would have far-reaching effects for mankind.

Q: What movies, documentaries or TV programs have you found interesting and/ or inspiring?
As a child, two television series greatly inspired me, (1) Carl Sagan’s “Cosmos,” a 13- hour series on cosmology, and (2) David Attenborough’s “Life on Earth,” a 13-hour series that surveyed all of life on the planet. David Attenborough continues at age 85 to make excellent programs on zoology, which I love to watch.

Q: In your spare time, what are you reading?
I read Richard Dawkins’ website once a week (http://richarddawkins.net/). It contains links and information about science, rational thought and evolution that I enjoy. I also like to read Sam Harris on religion and philosophy.

Q: How do your students inspire you?
My students come from very diverse backgrounds. Some of them have come from very humble backgrounds and have “pulled themselves up by their bootstraps,” as the saying goes, which is admirable.

Q: What advice do you have for students graduating in the next five years?
Read everything, not just in your area. Great ideas in computer science can come from entomology, economics—you name it. Also, don’t assume something is true just because someone famous said it. Several of my best papers have shown that well established “facts” are just not true.

Q: If you had unlimited resources and no constraints, what would you do to further your research?
My current project on insect classification requires me to collect a lot of data—there are 3,528 different kinds of mosquitoes. If I had unlimited resources and no constraints, I would be able to collect all this data. I actually think it may be possible to achieve this with little money! My idea is to crowdsource this.

Q: Do you have a personal hero?
I have several. Carl Sagan, David Attenborough and Richard Dawkins have always been scientific heroes. Bill Gates is a hero for deciding to give 99% of his money to charity.

Q: What is it about UCR that makes this a great place to be?
I love UCR because of the availability of expert colleagues. Anytime I need to talk to an expert on mosquito flight, the sounds made by ground squirrels, black holes or 15th century heraldic shields, they are less than a few hundred meters away.

Eamonn Keogh "Insect-borne diseases like malaria kill at least one million people a year. If we could accurately classify insects in the field in real time, we could plan much cheaper and effective intervention campaigns."

—Eamonn Keogh