To repository is used to host the source code used in the paper Regularized Text Logistic Regression: Key Words Detection and Sentiment Classification for Online Reviews. The code contains two parts, one for Restaurant dataset and one for Hotel dataset.

The Restaurant data is contained in the restaurant folder. For Hotel dataset, please refer to the reference in the paper.

For each experimentation, two files are used. The file working.R contains the high level workflow, with all the detailed implementation in the file helper.R.