 |
Muktha Ananda
(the artist formerly known
as AC Surendran)
|
Architect/Principal
Applied Researcher @XBOX
One Microsoft Way, Redmond WA 98052,
Phone: (425) 707-5018, Fax: (425) 936-7329,
email: Take my old name above, put a
dot between the names @MICROSOFT.com
|
About Me + My Interests
Currently I am working on gaming
intelligence for XBOX.
Previously I worked in
search, recommendations, social search, online advertisements,
knowledge bases, text mining and speech/speaker recognition, in the following
roles:
- Principal Dev Lead, Bing Interest Graphs
- Built systems to infer user interests & to recommend
things of interest.
- Principal Research Lead/Manager, Bing
Social Team
- Shipped semantic expansions in Bing Sidebar "Friends
Who Know"
- Co-founder, R&D Lead,
Knowledge Web team @Microsoft.
- Built
large
scale Knowledge Bases
- Shipped
Bing Finance - knowledge driven finance portal, helping people make
better decisions
- Designed & built push-model apps for restaurant
recommendation (iPhone), short video recommendation (tablet)
- Senior Applied Researcher,
Microsoft adCenter Labs
- Lead R&D teams responsible for Contextual
advertisement. Built
Video advertisement
prototypes, worked on Behavioral Targeting.
- Researcher,
Machine Learning@
Microsoft Research.
- Machine
intelligence applied to understanding, organizing,
tagging & discovery in email
- Text mining, growing topic models, text
categorization, clustering, etc
- Audio/Speech enhancement/noise suppression using
adaptive regression models for
teleconferencing.
- Researcher,
Bell Labs - large scale speech and
speaker recognition, microphone arrays
Some recent projects for which I managed the applied
research team:
Much of this work was on "finding connections that interest you". It is about helping people discover
things that interest and delight them & help them make better decisions. For
this, I have used data mining, machine learning, semantics & recommendations. I have
worked extensively in various aspects of this space:
- content understanding (semantics, knowledge bases, text
mining, query & document to topic mapping, content similarity)
- user understanding (interest graphs, interest modeling)
- context modeling (context genomes, appropriateness of content to context, etc)
- recommendation engines (movies, XBOX Live
games using collaborative & social filtering, restaurant, video recommendation)
- experiences (mobile, tablet & web experiences)
- Interest Graphs.
One simple example of what we have enabled is if you
search on Bing for "meat dress costume" we connect you in
the Sidebar to Facebook friends who like Lady Gaga.
|
|
- In the Knowledge Web team, we built the
very first large scale knowledge graph in Microsoft, which
has become the seed for Bing's own knowledge graph. On top
of this, we built experiences for discovery of knowledge.
The example here also shows ads that are semantic.
|
|
- Our first knowledge base was in personal
finance, and was the primary driver for Bing Finance. One
example of what we did was to provide alternate
recommendations when searching for a stock - other "higher
dividend paying", "top buy rated in the same sector",
"having similar PE ratio" etc. which helped people make
better stock picking decisions
|
|
- We also built recommendation apps
that
advanced a "push" paradigm (restaurant recommendation on a
mobile phone & video recommendation on tablet). They had
well-integrated social experience (Facebook friends), worked
on "context genomes" and "taste mash-ups"
|
|
Recent research community activities:
- PC Member, CIKM 2013 -
ACM Conference on Information & Knowledge Management, San Francisco, Oct
2013.
- PC Member, KDD 2012
- ACM Conference on Knowledge Discovery & Data Mining, Beijing, China,
August 2012.
- PC Member, WWW 2012 -
World Wide Web Conference, Lyon,
France, Apr 2012
- Workshop Chair, WWW 2010, Nineteenth International World
Wide Web Conference, Raleigh, NC, Apr 2010.
- Co-chair for
ADKDD '09 - The
3rd KDD Workshop on Data Mining and Audience Intelligence for Advertising,
June 28th 2009, Paris, France
- PC member,
Fifth Workshop on Ad
Auctions, July 6 2009, Stanford CA.
- PC member for
SIGIR 2009 Workshop on
Information Retrieval and Advertising (IRA 2009), Boston, July 23 2009.
- Associate Editor for the
IEEE Transactions on Speech, Language and Audio Processing (2004-2007)
- Co-chair for the First International Workshop on
Data Mining and Audience Intelligence for Advertisement (ADKDD'07) in
conjunction with KDD '07.
- PC member for TROA
2008 WWW
Workshop on Targeting and Ranking for Online Advertising, Beijing,
China, Apr 2008.
- PC member for
SIGIR Workshop on
Information Retrieval for Advertising (IRA 2008), Singapore, July 2008.
- Treasurer for
KDD 2008, Las Vegas, NV
- Co-chair for
ADKDD '08 - The
2nd KDD Workshop on Data Mining and Audience Intelligence for Advertising
- Senior PC member
IUI 2009 - International
Conference on Intelligent User Interfaces, Sanibel Island, FL, Feb 2009.
- PC Member, SDM09 -
Ninth SIAM International
Conference on Data Mining, Sparks, NV, Apr 2009.
- PC Member,
WWW 2009, Eighteenth International World Wide
Web Conference, Internet Monetization Track, Madrid, Apr 2009.
Selected Publications:
Advertising & Web:
- "Evolving Today's Web into the Knowledge
Web", A. C. Surendran, Tarek Najm & P. Vaddadi, Invited talk at the NIPS
2008 Workshop on Beyond Search: Computational Intelligence for the Web,
Whistler, Canada, Dec 2008.
- "Report
on the Second KDD Workshop on Data Mining for Advertising", D. Shen,
A. C. Surendran & Ying Li, SIGKDD Explorations, Vol. 10, Issue 2,
December 2008..
- "Learning from Multi-topic Web Documents
for Contextual Advertisement", Yi Zhang, A. C. Surendran, John C. Platt & Mukund Narasimhan,
KDD 2008, Las Vegas, Aug 2008.
- Best Application Paper Award Runner-up
- "Re-wiring the Ad Ecosystem",
A. C. Surendran & Tarek Najm, Keynote at IRA 2008, SIGIR Workshop on Information
Retrieval for Advertising, Singapore, July 2008.
- "Data Mining and Audience Intelligence for Advertising",
Y. Li, A. C. Surendran and
D. Shen, SIGKDD Explorations, Vol. 9, Issue 2, pp.
96-99, December 2007.
Semidefinite Programming:
Boosting:
Text Mining, Information Organization and
Retrieval:
-
"Incremental
Aspect Models for Mining Document Streams", A. C. Surendran and Suvrit
Sra, ECML/PKDD 2006, Berlin, Germany, Sep
2006.
-
SWISH: Semantic Analysis of Window Titles and Switching History", N.
Oliver, G. Smith, C. Thakkar and A. C. Surendran, Proceedings of Int. Conf. on Intelligent User Interfaces (IUI'06). Sydney. Australia. Jan
2006.
-
Automatic
Discovery of Personal Topics To Organize Email" A. C. Surendran, J. C.
Platt and E. Renshaw, Conference on Email and Anti-Spam, 21-22 July at Stanford University,
2005.
Acoustic Echo Suppression:
-
"Double-talk
Detection using Real-time Recurrent Learning", with M. Iqbal, and
others, to be presented to the International Workshop on Acoustic Echo and
Noise Control IWAENC '06, Paris, France, September 2006
-
"Regression-Based Residual Acoustic Echo Suppression",
A. S. Chhetri, Arun C. Surendran, J. W. Stokes and John
C. Platt, International Workshop on Acoustic Echo and Noise
Control IWAENC '05, Eindhoven, Netherlands, September 2005.
Speech Detection:
-
"Logistic Discriminative
Speech Detectors Using Posterior SNR", A. C. Surendran, S.
Sukittanon and J. C. Platt, ICASSP-2004, Montreal, Canada, May 2004.
- "Convolutional Networks for Speech Detection",
S. Sukittanon, A. C. Surendran, J. C. Platt and Chris J. C. Burges, Interspeech 2004, Korea, October 2004.
Bayesian Learning for speech recognition:
- “Transformation Based Bayesian Prediction for Adaptation
of HMMs”, A. C. Surendran and C.-H. Lee, Speech Communication (34),
pp. 159-174, April 2001.
- “Structural Bayesian Predictive
Adaptation of HMMs”, O. Siohan and A. C. Surendran, Workshop on Adaptation,
Sophia-Antipolis, France, August 2001
- “Hierarchical Bayes Approach to
Adapting Delta- and Delta-Delta Cepstra”, A. C. Surendran, ICASSP ’00, pp.
973-976, Istanbul, Turkey, June 2000.
Speaker Verification:
- “Speaker Authentication for Business and
National Security”, Invited talk at the IEEE Conference on Technologies for
Homeland Security, Boston, MA, April 2002.
- “Sequential Decisions for Faster and
More Flexible Verification”, A. C. Surendran, EUROSPEECH ’01, pp. 763-766,
Aalborg, Denmark, September 2001.
- “A Priori Threshold Selection
Mechanism for Fixed Vocabulary Speaker Verification Systems”, A. C.
Surendran and C.-H. Lee, ICSLP ’00, Beijing, China, September 2000.
- “Background Model Design for Flexible and Portable Speaker
Verification Systems”, O. Siohan, C.-H. Lee, A. C. Surendran and Q. Li, ICASSP '99, pp. 825-828, Phoenix, AZ, 1999.
Neural Networks for model adaptation and novel feature extraction:
- “Non-linear Compensation for Stochastic
Matching, A. C. Surendran, C.-H. Lee and M. Rahim, IEEE Transactions on
Speech and Audio Processing, Vol. 6, No. 7, pp. 643-655, Nov. 1996.
- “Towards Knowledge Based Features for
Large Vocabulary Automatic Speech Recognition”, B. Launay, O. Siohan, A. C.
Surendran and C.-H. Lee, ICASSP ’02, pp. 807-810, Orlando, May 2002.
Microphone arrays:
- “Inverse Problems in Microphone
Arrays”, A. C. Surendran, in the Handbook of Digital Signal Processing,
V. Madisetti, Editor, CRC Press/IEEE Publication, 1997.
- “Spatially Selective Sound Capture for
Speech and Audio Processing”, J. L. Flanagan, AC. Surendran and E. E. Jan,
Speech Communication, Vol. 13, pp. 207-222, 1993.
Other useless information:
Last updated: May 1st 2013.
