#Introduction This fold contains code and data to replicate the experiments for the binary classification task on Stanford Sentiment Treebank.
- binary_cnn: source code folder. binary_cnn/lexicon2/LexiconLambda.cc is the main code.
- data: data folder
- exp: scripts folder
#Usage ###Complie
cd binary_cnn
./cmake.sh
###Run
Assume you are in the binary_task folder.
cd exp
Train with SD-Lex
train_sd.sh
Train with SWN-Lex
train_swn.sh
###Accuracy You can use
grep Accuracy *.log
to find out the classification accuracy.