tingtang2 / yelp_cultural_analysis Goto Github PK

a topic modeling package written for our paper: Analyzing Cultural Assimilation through the Lens of Yelp Restaurant Reviews

Key utilities written in Cython for simple but fast computations

Code is adapted and heavily inspired by the lda PyPi library as well as Michael Paul's ccLDA implementation

To Cythonize:

python3 setup.py build_ext -b c3rlda/

Please download the dataset from the Yelp Open Dataset website.

Of those files, we use yelp_academic_dataset_review.json and yelp_academic_dataset_business.json.

To filter the dataset:

python3 preprocess_Yelp_dataset.py

To tokenize:

python3 preprocess_data.py -below 5

python3 run_{model}.py

python3 calc_NPMI_internal.py

python3 calc_palmetto_{model}.py