It is only the second day of the dashboard week and the pressure of daily project turnover has been both fun and demanding. The limitation on time means we have to make decisions on the contents of our dashboard relatively early on.

Making wrong decisions will lead to eventual backtracking and expenditure of valuable time, something I faced it with today’s challenge.

For Tuesday’s task, we were given the choice of either beer or wine datasets. I chose wine when I realized it had a very detailed description written by a professional Sommelier. I believed I could do some sentiment analysis using this description and quantify it.

I remember doing sentiment analysis using Twitter data during the Alteryx week with Andre and Phil. I believed I could repeat the workflow

Here is the sentiment macro I would be using for the sentiment analysis part.

 

Here is my workflow, I had used to complete the sentiment analysis.

 

However it was quickly evident that I could not do a thorough sentiment analysis as the description section written by Sommeliers did not contain any ‘sentiment’ words, instead, it only used adjectives to describe the taste, smell and structure of the wine.

Subsequently, I changed my dashboard idea and wanted to make a very visual work.

This dashboard is inspired by the wine list in restaurants. It is essentially a tool which can be used to compare wine bottles between two countries. To make this dashboard I went against few fundamental unspoken rules (sorry Andy):

  • I floated EVERYTHING
  • I made the dashboard scrollable

 

From this task, I have realized the importance of identifying and acknowledging the mistakes as early as possible.  I also learned how long it takes to fix the dashboard when everything is floated everything.

On that note, this is the end of Tuesday’s blog, I will see you tomorrow.