XLore_Feature

---

Code for feature part for paper:

---

Step

• Use Stanford parser for tagging and typed dependencies (ParserDemo)
• Calculate features (feature)

ParserDemo

• For English File

• ant main -Dargs='en input_file_name'
• ant main -Dargs='en input_file_name input_path'
• ant main -Dargs='en input_file_name input_path output_path'

• For Chinese File

• ant main -Dargs='zh input_file_name'
• ant main -Dargs='zh input_file_name input_path'
• ant main -Dargs='zh input_file_name input_path output_path'

• Default Input Path: /home/lmy/data/origin/
• Default Output Path: /home/lmy/data/parser/

Note: There may be mistakes when parsing. You can rerun the program. Unless you remove the parser result file, it wont reparse the sentence which has already been parsed.

Feature

####1.English

python headword_main.py The first 9 features python wordFrequency.py Add the 10, 11 features python add12thFeature.py

####2. Chinese

python getfeatures_ch.py The all 13 features

Change the path and file name in python file first