Recommending strains of cannabis based on sentiment of customer reviews

It’s no secret that the cannabis industry has huge market potential and realization. There are delivery services, high-end retail and subscription plans, and publicly traded companies, all of these are quaint little data collection agents. One big player in the game is a website and resource called Leafly – they hold a database of strains, dispensaries, and reviews. It is my goal to use my NLP skills and build a recommendation tool for strains, based off of the reviews customers leave on the strain. Leafly has thousands and thousands of reviews for over 3000 strains, so my sentiment data will be a huge addition to an already massive database.

This project combines my web scraping skills with my NLP scripts (a classic for me if I must admit) to create a new tool forĀ cannabis enthusiasts – one that recommends new strains to you based on how you feel about your favorite strains.

3420
Strains
100000
Reviews

*New strains and reviews to be added on idk, a monthly schedule? How often are news strains made?

Strains matched to your taste and feeling

Go follow the project on Github!

Time to find your next favorite strain and leave a review today

Hey you know what, I’ll even help you with that! Leafly, please feel free to invite me to the Seattle office and I would be honored to explain the process and code in person, with a live demo of course.