Yamini  Kannan

 

Email: yamini [at] cs.berkeley.edu

 

Computer Science Division

University of California at Berkeley

566, Soda Hall,

Berkeley, CA 94720.

 

 

Bio:

I am a first-year graduate student in the Electrical Engineering and Computer Sciences Department at UC Berkeley.

My research interests include Programming Languages, Program Verification and Analysis, Software Engineering, Formal Methods.

Formerly, I worked in Microsoft Research India in Bangalore as a Research Software Design Engineer. I was a member of Advanced Prototyping and Development group. I was also an affiliate member of Rigorous Software Engineering group and Digital Geographics group.

Prior to that, I did my undergraduate studies in Computer Science and Biological Sciences at Birla Institute of Technology & Sciences, Pilani.

For more details, refer my resume: pdf/ text

Check out my brother’s page.

 

Publications:

o   Bhargav S. Gulavani, Thomas A. Henzinger, Yamini Kannan, Aditya V. Nori and Sriram K. Rajamani. Synergy: A New Algorithm for Property Checking. Proceedings of the 14th Annual Symposium on Foundations of Software Engineering (FSE), ACM Press, November 2006 (ACM-SIGSOFT Distinguished Paper).

 

o   Yamini Kannan and Koushik Sen. “Universal Symbolic Execution and its Application to Likely Data Structure Invariant Generation”. Submitted to International Symposium on Software Testing and Analysis (ISSTA) 2008.

 

Projects:

o   Universal Symbolic Execution and its Application to Likely Data Structure Invariant Generation. (With Koushik Sen, UC Berkeley)

o    Proposes a new algorithm for inference of likely invariants of data structures. Makes use of a novel variant of symbolic execution called Universal Symbolic Execution, that is intertwined with concrete execution to generate invariants that are highly relevant to the program.

 

o   Target-oriented binary analysis (With Dawn Song, UC Berkeley)

o    Proposed and implemented a target–oriented path exploration algorithm that explores program paths in a lazy, modular fashion. Inter-procedural analysis is enhanced by using a context–insensitive approach.

o    Project write-up

 

o   Yogi (With Sriram Rajamani, Aditya Nori at MSR India)

o    Yogi is a research project within the RSE group on property checking. This involves a new algorithm that combines static analysis and testing techniques in a systematic manner to prove safety properties in programs, and is implemented to work over program binaries.

o    Download FSE presentation from here: (ppt) (pdf).

o    Poster on Yogi presented in TechVista 2007 (Annual Research Symposium organized by MSR, India).

 

o   Virtual India (With Joseph Joy , MSR India)

o    Virtual India is a project within the ADP group, done in close collaboration with the TerraServer and Virtual Earth teams, to build a platform that allows end users to collaborate over the Internet to augment existing maps in rich ways.

o    We released a publicly accessible prototype version in January 2006: http://virtualindia.msresearch.in/  (Zoom into Bangalore for interesting stuff)

 

Courses:

o   CS294-25: Current Berkeley Research in Programming Systems (Fall 2007)

o   CS294-24: Privacy and Security Enhancing Technologies (Fall 2007)

o   CS294: Practical Machine Learning (Spring 2008)

o   CS270: Combinatorial Algorithms and Data Structures (Spring 2008)

o   CS263: Design and Analysis of Programming Languages (Spring 2008)