Office: Gates-Hillman Complex 8103.
Phone 412-268-7678. Fax 412-268-5576.
Secretary: Christina Melucci,
GHC 6509, 412-268-1593.
Mailing address: Carnegie Mellon / Computer Science, 5000 Forbes Ave.,
Pgh PA 15213
Hi! My research interests are in:
- Forecasting Epidemics:
The long term vision of our DELPHI
research group is to make epidemiological forecasting as universally
accepted and useful as weather forecasting is today. As was the case with weather
forecasting, this will likely take several decades. In the shorter term, we select high
value epidemiological forecasting targets (currently Influenza and
Dengue); create baseline forecasting methods for them; establish metrics
for measuring and tracking forecasting accuracy; estimate the limits of forecastability for each target; and identify new sources
of data that could be helpful to the forecasting goal. We are part of the
multi-university MIDAS research
- Information and
Communication Technologies for Development (ICT4D),
and specifically Spoken Language Technologies for Development (SLT4D), which
is the term we coined for our own subfield of ICT4D: finding ways to use
spoken language technologies (like automatic speech recognition, speech
synthesis, and human-machine dialog systems) to aid socio-economic
development around the world.
Our current project, Polly, uses telephone-based viral
entertainment to reach low-literate people in Pakistan and India, familiarizing
them with speech interfaces and then introducing them to development-related
services. First deployed in Lahore in
May 2012, Polly reached over 165,000 users all over Pakistan and fielded over
2.5 million phone calls in 8 months. In
2013 we launched Polly in Bangalore, India, and it ended up spreading virally
to West Bengal, New Delhi and other areas of India. Since March 2015, we have been deploying Polly in Guinea, for
person-to-person spreading of approved Public Health messages about Ebola in
many languages, in collaboration with the US embassy in Conakry. Try a live
demonstration of our current system.
A previous project, HealthLine,
investigated the use of a telephone-based automated dialog system for access to
healthcare information by low-literate community health workers in Pakistan.
- How, and to what
extent, can the evolution of infectious diseases like Influenza be
- How, and to what
extent, is the evolution of viral disease like Influenza affected by
public health interventions such as vaccination, antiviral drug use,
school closures, travel restrictions, etc.
We model the spread of epidemics in the population as well
as the evolution of the virus itself, such as changes in its virulence,
pathogenicity, drug resistance, or antigenicity (immune escape).
Students: David Farrow (CompBio Dept, viral evolution +
Epi-forecasting), Logan Brooks (CSD, Epi-forecasting).
Ali Raza (LTI,
ICT4D), Chuang Wu
(CompBio Dept, viral
genotype-phenotype mapping), Jahanzeb Sherwani (CSD, ICT4D), Yong Lu (CSD, CompBio), Dan Bohus (CSD, dialog
systems), Stefanie Tomko
speech communication), Jerry
(Xiaojin) Zhu (LTI, MLD, semi-supervised learning), Lin
Chase (RI, speech recognition).
Past Post-docs: Andy
Walsh (computational virology), Xiaojin Wang
(machine learning), Stan F. Chen (language
modeling), Pierre DuPont (language
My favorite quotes.