NLP

Collect Event Datas from all over US

1. Find mapping from eventbrite category id to name
   1. Eventbrite Categories Here is the official list of Event Categories used by Eventbrite.
   2. Run extract_categories.py to create the mapping from event id to name.
2. Scrape from Eventbrite API
   1. Run response_formatter.py and collect event data in batch. Only store each event's description, title and category name (from our eventid2name mapping obtained in last step).
3. Upload Formatted Event Data to Firestore
   1. Obtain firebase service key json file (Do not put in Github, or any public place)
   2. Run upload2firebase.py to upload event data to "unlabeled" collection in firestore.

Tagging is done using the tagging service

Trained classification model and predict category & preference tags for our events

4. Download the labeled events data from firestore
   1. Run downloadTrainSet.py to download labeled events from "labeledDescriptions" collection in firestore.
5. Train the text classification model
   1. Run model_category.ipynb
   2. Run model_preference.ipynb
6. Do the inference on unlabeled data
   1. Run Inference.ipynb