See "Report.docx" for details of our goal to extract different types of pizza from yelp reviews. You may also directly open the each ipython file to see each step of our process documented and the associated code.
To run our project perform the following steps:
-
Install install DDLite ( https://github.com/HazyResearch/ddlite )
-
Install Jupyter ( https://ipython.org/ )
-
Download Yelp Academic Dataset ( https://www.yelp.com/dataset_challenge/dataset )
-
Extract dataset json files to the subfolder "yelp_data"
-
Download "Yelp_Tagger_Learning.ipynb" and "YelpTagger_Extraction.ipynb" from the repository
-
Run "jupyter-notebook YelpTagger_Extraction.ipynb" to perform candidate extraction
-
Run "jupyter-notebook YelpTagger_Learning.ipynb" to perform learning and evalutation