With the goal of making a data-driven choice for our next hamburger lunch, we gathered information about restaurants in Amsterdam from a review website. The data contains reviews, ratings, and pictures. In this talk we will examine and combine these three different pieces of information to make our decision.

In the talk, I will demonstrate how to collect data from the internet using web scraping package Scrapy. To give it a bit more of an engineering feel, we will run the scraper from a Docker container on Google cloud and push the collected data into BigQuery. I will then show how you can easily retrieve data from BigQuery with Pandas, and we will analyse the ratings. Sentiment in review texts is quite easily extracted using ElasticSearch. And the collected images are send to the Google Vision API in order to classify if there's a nice and juicy burger on it. Although all this might sound like an overkill of tools, the point here is to experiment and learn. There are just too many toys out there to play with!