- Python 3.6
- Install these
pip install --upgrade --user git+git://github.com/zhijing-jin/efficiency.git
# many other packages as mentioned in code/parser
- CoreNLP Package
- Download a java package onto the server, Corenlp
- Then install the python api
- 大致就是先download java的package，然后用python的api跑一下，其中需要开server的port 9000。
- Install NLTK with
sudo pip install -U nltk
- Play with CoreNLP parser by
cd code/parse
python corenlp.py
- Run Story_Matching examples by
cd code/parse
python start.py -file success
-
Run like-analogy model by
- Start up Stanford CoreNLP server with minimum 4 GB of memory (can customise)
cd stanford-corenlp-full-2018-02-27 java -mx4g -cp "*" edu.stanford.nlp.pipeline.StanfordCoreNLPServer-
Activate virtual environment with
source activate python3 -
Run the script with
python like_analogy.py -
Input query in format of "I like ___." where ___ can be "wolves", "cats", etc.