rec_engine_stream

usage

poetry install

poetry run python3 -m rec_engine.run_read_data

Google local review

These datasets contain reviews about businesses from Google Local (Google Maps). Data includes geographic information for each business as well as reviews. Download the full datasets from https://cseweb.ucsd.edu/~jmcauley/datasets.html#google_local.

Users are indexed by gPlusUserId and items(place) are by gPlusPlaceId. User and item features are contained in users.clean.json and places.clean.json. For example, for each gPlusUserId, we know his/her jobs, education, and currentPlace.

To treat the data as a streaming data, we let the review time reviewTime be the time variable t.

Papers using the data

Two papers are saved in the paper folder.

Useful links

https://github.com/blei-lab/context-selection-embedding

https://www.slideshare.net/justinbasilico/recent-trends-in-personalization-at-netflix?next_slideshow=1