Morph Resolution README
1) Install word2vec
	a) In word2vec directory, run 'make' to compile word2vec tool.
	b) You can run the demo scripts: ./demo-word.sh and ./demo-phrases.sh to test if installed successfully.

2) Preprocessing tweets
	a) Concatenate all Chinese tweets into one file.
	b) Run any Chinese segmentation tool on Weibo dataset. One light Chinese segmenter is Jieba which is a python based segmenter. You can find it at: https://github.com/fxsjy/jieba

3) Train models
	Run train_model function in morph_resolution.py to obtain a model on Chinese tweets. Make sure to change the path.

4) Morph resolution
	Use function candidate_ranking to get answers for each query.

4) Evaluate model
	Use do_evaluation to evaluate the result according to gold standard.


(If you have any question, please email zhangb8@rpi.edu)