The Project required conceptualisation and implementation of a framework using ActionScript to draw Graphs on the fly.  Then subsequently, the framework to draw graphs, was modified for three different projects on Visualisation of Graph Algorithms to help beginners learn Graph Algorithms.

  • Graph Traversal
  • Shortest Paths in Graphs
  • Minimum Spanning Trees

                         Mad Android is  a First Person Shooting Game written in C++ using openGL for rendering, and guess what,  its written right from scratch (except for the rendering part, which I let openGl handle). It started when the Proff. in Graphics course, wanted us all to make a 2D projectile game, and when is that ?

 
                       As a part of my Research Assistantship @ CCNSB, IIIT Hyderabad, I was required to build interactive applications in Flash, which simulate experimental conditions and help students to learn a particular experimental topic thoroughly, via an almost real life experience.  The user is expected to interact with the application at each and every level of the experiment, and thus the user understands the how and whys of all the steps. 

Wordnet can get really messy sometimes to visualise. And I dont like messy things. I cooked up this wordnet visualiser for the Hindi Wordnet released by Prof. Pushpak Bhattacharyya and his team at IIT-Bombay.

They have released the Wordnet and the API at : http://www.cfilt.iitb.ac.in/wordnet/webhwn/

 

The Wordnet visualiser is hosted at : http://web.iiit.ac.in/~sharada.mohanty/wordnet/d3/

 

 

IIIT Hyderabad calls itself a research institute. With Research comes publications in conferences, with conferences come travel and with travel come travel grants :) ! But how exactly are the travel grants distributed ? A student gets a publication in a major conference, he sends in a mail, with a whole discourse on why he should be given a grant. Now the professor forwards it to the travel Grants committee, and the poor travel Grants Commiittee has to sit down, and read all such application (read essays) and filter out the relevant information, and then prioritize based on the available funds, and then finally say the account Department...All ryt give him this much money . Poor them !!

  • www.SanketSharma.com

              Sanket Sharma, a traveller and photographer, wanted a website for all his travelogues, photo galleries, tour informations, and most importantly a proper admin interface that would let him add , remove, modify any content without writing one single HTML tag, or typing in one single command into the terminal. And, being a photography website, he was also very particular about the aesthetics of the final website,
     
  • web.iiit.ac.in/~parliament


             A body like Student Parliament, needs quite some workflow management. There are Issues, which some one raises, There are issues which need discussion, all the discussions of parliamented has to be at a place where everyone and anyone can access them anytime, People need a place to Vote, People need a forum to talk. so many problems....being a parliamentarian and a web designer, I had to come up with something ;) !! and here it was.....the Student Parliament Website.

"John hit Mary. Why in the world did he have to hit Mary ?"

              If I have a sentence like this, everything is fine when processing with a machine, except when we come across the "he" . Who does "he" refer to ? John ? Mary ? The answer is quite intuitive for us humans, but quite "out-of-the-world" for a machine. 

              This very problem is popularly known in the Computational Linguistics community as the problem of Anaphora Resolution. In a layman's terms it is the problem of resolving what a pronoun or a noun phrase refers to.

The language used on twitter is incoherentunstructured and far different than something the modern day parsers like stanford parser or berkleys parser were used to. So a proper linguistic analysis was a far cry.  Before attempting to analyse the pro-drop features in the language, another problem needed to be solved, that being How do we enable our current linguistic toolset to work the same way it  does with a normal well structured english sentence, any english teacher would write on the blackboard. Another was, to try and make the language  compatible with the modern day linguistic toolkits, as much as we can. Methodology I used the twitter api, to extract some number of tweets based on keywords like :ban, awesome, cool, god, corruption, exams., and then analyse them and try to filter out noises and hance make them more compatible with our Computational Linguitics tool kits like NLTK library, Stanford Parser, Berkley Parser, Brills Tagger, and the likes.

Skillset

  • Programming Languages

    •  C/C++
    •  Python
    •  Perl (Basic)
    •  Java (Basic)
    • Shell Scripting , BASH 
  • Web Languages

    •  HTML
    •  CSS
    •   XML
    •    RDF (Basic)
  • Serverside Scripting
    •  PHP
  • Clientside Scripting
    •  Javascript
    •  AJAX
    •  jQuery
    •  JSON 
  • Query Languages
    • SQL
    • SPARQL (Basic)
  • Web Frameworks
    • WEB2PY
  • Content Management Systems
    •  Drupal
    •  Wordpress (Basic)
    •  Joomla (Basic)
    •  Silverstripe (Basic)
  • Learning Management Systems
    •  Moodle (Basic)
  • Operating Systems
    • LINUX
    •  Windows
    • MAC OS X
  • Other
    •  Adobe Flash CS5
    •  Actionscript 3.0
    • BigBlueButton