User:Gang Chen

About me

Name: Gang Chen

Email: pkuchengang@gmail.com

IRC: Gang

SourceForge: elephantgcc

GitHub Repo: https://github.com/elephantgcc

GSOC 2013

I'm working with Apertium for the GSoC 2013, on the project "Sliding Window Part of Speech Tagger for Apertium".

my proposal is here: Proposal

svn repo

https://svn.code.sf.net/p/apertium/svn/branches/apertium-swpost/apertium

LSW tagger: Current Progress

http://wiki.apertium.org/w/index.php?title=User:Gang_Chen/GSoC_2013_Progress

Useful tools

Wikipedia extractor

target

This tool extracts main text from Wikipedia, producing a text corpus, which is useful for part of speech tagging, language model training, etc.

tool

http://code.google.com/p/natural-language-qa/source/browse/MakeCorpus/WikiExtractor.py

usage

1. Get the script.

  http://code.google.com/p/natural-language-qa/source/browse/MakeCorpus/WikiExtractor.py

2. Download the Wikipedia dump file here:

   http://dumps.wikimedia.org/backup-index.html

Take Chinese as an example, download the file zhwiki-20130625-pages-articles.xml.bz2 on this page http://dumps.wikimedia.org/zhwiki/20130625/. Alternatively, we can download the latest version on this page, http://dumps.wikimedia.org/zhwiki/lastest/

3. Use the script.

mkdir output
bzcat zhwiki-20130625-pages-articles.xml.bz2 | ./WikiExtractor -o output

Optionally, we can use

"-c" for compression for saving disk space, and

"-b" for setting specified bytes per output file.

More information please type "./WikiExtractor --help".

Ok, let's have a cup of tea and come back an hour later.

The output should be output/AA/wikiXX, where wikiXX are the extracted texts.

4. clean up "<>" tags

We are one step from the final text corpus, because there links in wikiXX files. Let's use the following tiny script to filter out "<>" tags.

#! /usr/bin/python
# -*- coding:utf-8 -*-
import sys
import re
regex = re.compile(ur"<.*?>")
line = sys.stdin.readline()
while line != "":
	line = regex.sub("", line)[ : -1]
	print line
	line = sys.stdin.readline()

5. done :)