Difference between revisions of "User:Akoul"
From Cohen Courses
Jump to navigationJump to search| Line 22: | Line 22: | ||
== Brainstorming Idea == | == Brainstorming Idea == | ||
| − | + | ||
| − | * Automatic extraction of answering patterns for Question Answering | + | === Extracting Facts from Wikipedia Text to generate a Fact relation hierarchy === |
| − | + | * Corpus : Wikipedia XML Dump | |
| + | === Automatic extraction of answering patterns for Question Answering === | ||
| + | * Corpus : Yahoo Research Webscope Dataset for Yahoo Answers | ||
Revision as of 14:46, 6 September 2011
Contents
Anirudh Koul
Who am I and Why I am here
Hello there, I am Koul (Pronounced 'Cool') from Canada. I am a Masters students in the Very Large Information Systems program (between LTI & ISR).
Before joining CMU, I worked 4 years with Yahoo as a Research Engineer. I am here to gauge what new technologies can be build by combining NLP and Machine Learning.
Superpowers
- Can function sanely in -40 Degree weather, and still keep continuing experimentation.
- Have crashed IBM Watson : Proof
- Have developed 3 Question Answering systems, once personally, then under Dr Eric Nyberg and later under Dr Noah Smith.
Random Notes
- Once ran a triathlon with a broken toe bone, just to look better than an MTV VJ right behind me, who was being captured on camera all along.
- Have a photo site with over a million visitors, 2000 sites currently using my Creative Commons licensed photos, including New York Times, Wikipedia, Yahoo Homepages, ABC News, ReadWriteWeb, etc. Photogallery
Brainstorming Idea
Extracting Facts from Wikipedia Text to generate a Fact relation hierarchy
- Corpus : Wikipedia XML Dump
Automatic extraction of answering patterns for Question Answering
- Corpus : Yahoo Research Webscope Dataset for Yahoo Answers
