Today’s challenge is about web scrapping and Tableau server.
The data is the Universities rank from World University Rankings 2021-22 | Global 2000 List | CWUR, what we have to do is webscrap the table in Alteryx, to achieve this, the Regex skill is quite important, I would highly recommend our Regex master Thang’s blog Alteryx Regex – From ‘zero’ to ‘hero’ – The Data School Australia, as it has a very complete guide of Regex.
To retrieve the web source code, we simply need to utilise the Alteryx download tool with the website url as the input, which is fairly simple. Building the Regex function to parse the data is the most difficult part. RegExr: Learn, Build, & Test RegEx is a website I recommend for building your Regex function. Basically, I copy and paste some source code into the text box first, and then we can check whether or not what we’ve done is right.
For this situation, I ended up writing a really long function. I’m sure there’s a simple solution, but it was a nice exercise for me because I wasn’t very comfortable with Regex previously.
After testing the Regex function, we just need to put it in Alteryx. Another thing to note is that before parsing the data, we must first Tokenize the Universities’ record into a row, which is a rather straightforward procedure to implement. I added some extra contexts because there aren’t many things to say with the current data. My theory is that we all know a school’s rating is essential, but does it truly correspond with job or salary? So I decided to focus mainly on Australian schools and research the employment rate. The website I have found is Home (qilt.edu.au) and they have many comprehensive reports after conducting survey on graduates.
Dashboard design and insights
The process of creating a dashboard on a server differs significantly from that of creating one on a desktop. My overall server experience was not particularly pleasant. The server has many limitations, and the activities may not always work. The dashboard I created on the server looks like this, but there are too many restrictions on adding annotations, so I switched to desktop and later published the workbook to the server.
I have built two versions of dashboard on the desktop. The design idea is from Andy Kriebel’s makeover Monday dashboard Profile – andy.kriebel | Tableau Public, he has a wide range of different dashboards and they will definitely give you the inspiration on building your data visualizations.
Basically, I was comparing school rankings based on a variety of factors, including CWUR’s score, full-time employment rate, labour force participation rate, median wage, and overall employment rate.
The conclusion is that the CWUR rank does not necessarily affect the employment rate and salary at Australian institutions, as we see radically different results when we alter the measurements. So, does the school rank matter in Australia? I don’t believe it does, based on the employment rate and wage.
And that is my challenge 4 project! Hopefully you will like it.
Feel free to leave any comments or feedbacks.