You do understand correctly!
The main idea about NLP components is with POS tagger as an example:
1. a fall back system that does unsupervised POS tagging. 2. the ability to plug in an existing POS tagger as these become available for specific languages.
I would as supervisor would recommend working with 3 languages. English, Hebrew, and the GSOC native language.
If we could get QA from other native speakers we would incorporate them into the workflow.
I think that by using a deletion/reversion based heuristic we may also be able to make a spam corpus to boost the accuracy of the corpuses.
Operation Manager E-mail: oren@romai-horizon.com Mobil: +36 30 866 6706
Római Horizon Kft. H-1039 Budapest Királyok útja 291. D. ép. fszt. 2. Tel: +36 1 492 1492 Fax: +36 1 266 5529
-----Original Message----- From: wikitech-l-bounces@lists.wikimedia.org [mailto:wikitech-l-bounces@lists.wikimedia.org] On Behalf Of Amir E. Aharoni Sent: Tuesday, April 03, 2012 10:19 PM To: Wikimedia developers Subject: Re: [Wikitech-l] GSoC 2012: Proposal-Wikipedia Corpus Tools
2012/4/3 karthik prasad karthikprasad008@gmail.com:
Hello, I am a GSoC aspirant and have compiled a proposal for one of the project ideas - Wikipedia Corpus Tools. [Mentor : Oren Bochman] I would sincerely appreciate if you could kindly go through it and suggest corrections/additions so that I can settle with a coherent proposal.
Link to my proposal : https://www.mediawiki.org/wiki/User:Karthikprasad/gsoc2012proposal
Nice, but why only English?
If i understand the proposal correctly, this project is supposed to be able to work with almost any language with very little effort.
-- Amir Elisha Aharoni · אָמִיר אֱלִישָׁע אַהֲרוֹנִי http://aharoni.wordpress.com “We're living in pieces, I want to live in peace.” – T. Moore
_______________________________________________ Wikitech-l mailing list Wikitech-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikitech-l