Priya Goyal



Department of Mathematics and Statistics

Indian Institute of Technology, Kanpur, India


 

[Home]     [Projects]     [Achievements]     [Resume]     [Contact]    


Here is the list of relevant projects I have done so far...

    Internships


  • Development of software to perform threshold based alerting and predictive analysis for SDLC build cloud
    (Software Developer Internship, Goldman Sachs, New York City, 2014)  

    The firm-approved SDLC cloud hosts thousands of product build events on a daily basis. In order to manage the expanding build cloud more efficiently, we needed a data driven automated mechanism to manage and report on the cloud utilization and capacity. To make the build cloud self-aware and dynamic, I developed software that looks for trends and threshold violations in the build analytics data. The engine reads user submitted queries, processes them on real-time build data and gives the trends/ violations.



  • Data Leakage: Collusion Graph for determining usage of Beacons and their providers
    (Software Developer Internship, Yahoo!, 2013)   [paper]  

    Developed a visualization tool for tracking the usage of beacons & their providers in online display ads to determine whether they are coming from the authorized/approved vendor or not. Paper presented at the annual internal technical conference of Yahoo! held in Santa Clara, USA 2014. Also developed a framework for integrating third party client for scanning ads served through Right Media Exchange, Yahoo!



  • Research projects


  • Peer-to-peer insult detection in online communities
    (with Prof. Amitabha Mukerjee, Department of Computer Science, IIT Kanpur, 2013)     [ppt]     [report]  

    We worked on building a classifier to detect the comments that are considered as insulting to other participants of blog/forum conversation. We used n-grams, skip grams and special words like ‘you’ to build vector model. We presented two new hypotheses to construct more feature and used tools in supervised learning like logistic regression and SVM for training and testing.



  • Intelligent Tutor System: Making Local and Global context motivated Interventions
    (with Prof. Carolyn P. Rosé, Department of Computer Science, Carnegie Mellon University, USA, 2012)    [ppt]    

    We worked on implementing an interactive agent moderated chat system, to assist a team collaborating on a common task. The interactive agent was capable of detecting confusion or distraction among team members and intervene appropriately, to keep the team in check, sometimes even with member specific interventions.



  • Algorithms for high order representation of complex 3-D surfaces
    (with Prof. Akash Anand, Department of Mathematics and Statistics, IIT Kanpur, 2012)   [ppt]   [report]  

    We worked on designing algorithms for fast and automatic high-order representation of complex 3-D surfaces via Fourier Continuation technique. Earlier, the process of surface generation was manual and required a lot of human intervention particularly in the selection of intermediate good projection surfaces. We successfully designed algorithms to automatically find good projection surfaces and gave an efficient algorithm for projection of given point cloud on it





"Genius is 1% talent and 99% percent hard work..."   - Albert Einstein