Oct 1, 2016 : Tata Consultancy Services (TCS) - Research Interaction Day
Tata Consultancy Services (TCS) Research Interaction Day will be held on Oct 1, 2016, Saturday, ICSR Hall 2, IIT Madras. The event will facilitate interaction between Research Scholars funded by TCS PhD scholarship, researchers from TCS and faculty members at IIT Madras. See the schedule
for more details about the program. See also the TCS PhD Scholar program flyer
Poster Presentation :
|Ditty Mathew||Creating Pedagogic View of Content from Encyclopedic Resources|
|Geethu Miriam Jacob||Video Stabilization and Moving Object Segmentation on Jittery Videos|
|Jithin Vechery||CGQ : Context-Aware Fast Graph Search in Large Networks|
|Jyothi Vedurada||Identifying Refactoring Opportunities for Replacing Type Code with Subclass and State|
|Sowmya Sundaram||Solving Simple Arithmetic Story Word Problems|
Here are the details of the talk(s) :
Speaker: Dr. Sayan Ranu
Title: Trajectory Analytics
Abstract: Recent advances in location tracking technologies has led to an explosion in the availability of trajectory data. However, the way trajectories are collected today has undergone a dramatic shift, which has rendered existing techniques ineffective. Instead of careful monitoring through accurate GPS devices, majority of the trajectory data today is generated as a by-product of another service. Examples of such services include call detail records from telecom operators, social network check-ins, credit card transactions, geo-tagged photo albums, and smart phone messaging apps. Consequently, there is no quality control and these datasets are noisy, inconsistently sampled, multi-modal, and highly dynamic. Due to this phenomenon, we reside in a data-rich but informationally poor world. In this talk, we will discuss the research efforts in my group that seeks to bridge this gap.
Speaker : Dr. Chester Rebeiro
Title : Automatic Detection of Side Channel Vulnerabilities in System Software
Abstract: The discovery of side-channel vulnerabilities in system software and their exploitation is a laborious manual task. As a
result side-channel attacks are difficult to perform and often reported as significant achievements. The apparent rarity and
difficulty of side-channel attacks combined with reservations that software developers have towards treating side-channel
vulnerabilities as bugs - since vulnerabilities do not result in incorrect behavior - has an undesirable result of systems
software shipping with many side-channel vulnerabilities.
In this talk we present our ongoing work and hurdles in developing a framework to automatically locate
side-channel vulnerabilities in system software.