Data Webscraping for Mechanical Keyboard
The skills the author demoed here can be learned through taking Data Science with Machine Learning bootcamp with NYC Data Science Academy.
Introduction:
I am a mechanical keyboard enthusiastic and I often introduce mechanical keyboards to my friends. However, the mechanical keyboard market is so diverse and filled with different data, so finding the perfect keyboard just for the person is quite a daunting task.
So I thought, why don’t I find a place with all kinds of keyboards available, use my knowledge of mechanical keyboards, categorize them and make a keyboard picker/recommendation system?
Scraping the data:
I chose “mechanicalkeyboards.com” as the website to scrape. Based in Tennessee, “mechanicalkeyboards.com” is one of the largest mechanical keyboard vendors in the United States. It offers not only pre-built complete keyboards but also parts and accessories for mechanical keyboards.
“Mechanicalkeyboards.com” has a simple and efficient architecture. Since it does not use REST API to retrieve the product listings, I decided to use Scrapy in Python to scrape the website.
I collected 797 listings from the online shop, with their:
- Id
- Product name
- Brand
- Model
- Mechanical switch type
- Price
- Material of Keycap
- Keycap legend color
- Keycap printing method
- Frame color
- Interface
- Dimension
- Weight
- Average rating
- Number of reviews
Using the information on hand, I made a shiny webApp for picking a mechanical keyboard. Also, check out my the GitHub repo for the code.
Some interesting data finds
Top 5 highest rated brands are:
Top 5 most reviewed brands are:
Notice these five are the all-star brands from the website.
For mechanical keyboards, the mechanical switches for the keys are the defining component. We see that the most offered mechanical switch brand is Cherry, which is the industry leader in producing the mechanical keyboard switches.
However, we see that Cherry mechanical switches are not the most expensive switches that the website offers. The electrostatic switches offered by Topre has the highest average price. We also see KBT, Gateron, and Kailh, which are Cherry knockoffs have generally lower average price.
While keyboards using no lighting or white lighting has the highest average price, keyboards using multi-colored lighting has the lowest average price.
And while keyboards using blue lighting has the highest overall rating, keyboards using pink lighting has the lowest overall rating.
Future works
Some NLP analysis using product reviews, and incorporate them into the shiny app.
About Author
