This package contains the CRF Baseline and LSTM-RNN implementation for the paper named "Fine-grained Opinion Mining with Recurrent Neural Networks and Word Embeddings" published in EMNLP2015, Lisboa, Portugal.

Prerequisite

To run the scripts, the datasets and some open source tools need to be downloaded: (1) Datasets from SemEval-2014 Task 4: http://alt.qcri.org/semeval2014/task4/index.php?id=data-and-tools and put in the evaluation folder. (2) word2vec and Google News Embeddings: https://code.google.com/p/word2vec/ (3) Amazon Reviews: https://snap.stanford.edu/data/web-Amazon.html (4) SENNA Embeddings: http://ronan.collobert.com/senna/ Note that the Embeddings should be put in the embeddings folder.

Example Commands

Some example commands are as follows: (1) bash rnn-batch.sh Senna (2) bash cv-batch.sh laptop Senna 50 50

Credits

(1) We used the tool CRFsuite for the CRF baseline, please refer to http://www.chokkan.org/software/crfsuite/. (2) For Elman-RNN and Jordan-RNN implementation, please refer to https://github.com/mesnilgr/is13.