GitHub - rrenaud/election-market-news-predict: My final class project for NLP.

Branches Tags

Name		Name	Last commit message	Last commit date
Latest commit History 31 Commits
intrade_data		intrade_data
iowa_data		iowa_data
labelled_sentences		labelled_sentences
mult_sentences		mult_sentences
.gitignore		.gitignore
README		README
boss_scrape.py		boss_scrape.py
cand_sentences.py		cand_sentences.py
cand_sentences_test.py		cand_sentences_test.py
candidate_info.py		candidate_info.py
date_util.py		date_util.py
english.stop		english.stop
fetch_search_links.py		fetch_search_links.py
intrade_data_parser.py		intrade_data_parser.py
model_debug_info.py		model_debug_info.py
model_debug_server.py		model_debug_server.py
predict_markets.py		predict_markets.py
scrape_intrade.py		scrape_intrade.py
sentence_labeller.py		sentence_labeller.py
textualize_html.py		textualize_html.py
vw_learner.py		vw_learner.py

Repository files navigation

This is an implementation of the research below.

http://www.cs.columbia.edu/~kathy/cs4701/documents/ReadingtheMarkets-ai.pdf
http://www.seas.upenn.edu/~cse400/CSE400_2006_2007/GilderLerman/poster.pdf
http://www.cs.jhu.edu/~mdredze/publications/markets_coling08.pdf

Software requirements:

market prediction:
         
numpy
vw - https://github.com/JohnLangford/vowpal_wabbit/wiki

scraping:

To run the searches, you need a yahoo BOSS account.

nltk - http://www.nltk.org/
w3m 

model debug/sentence label server:
webpy