The Data For This Dashboard
The data for today’s dashboard is an API containing details on the 480,000+ artworks held within the Metropolitan Museum of Art. It contains a plethora of details surrounding the different artworks, such as classification, medium, origin and end date, etc.
Data Collection and Cleaning
To create the data from the API call, Alteryx was utilised to collect and clean the data. The workflow itself is fairly simplistic, as the data is stored in a very convenient way that is easy to use in Tableau straight away.
One limitation within the data collection was the time it took to return the results, taking half an hour or so to return the 7k artworks used in the dashboard.
Introducing My Topic
The subsection I chose to hone my topic on was Medieval artworks. My original idea was to focus on comparing periods within the Middle Ages since it spans such a long stretch of time. Limitations within the data prevented me from doing this, however, so I decided to focus on a specific field in the data: ishighlight. This field indicates if an artwork is a highlight (popular) within the museum, so I attempted to see if there was a correlation between these artworks at all.
The Dashboard Overview
The overview serves as a good introduction to the database and dashboard as a whole, easing the viewer into the topic with some interesting statistics relevant to the highlight and non-highlight works.
Data Analysis
This section contained the major analysis of my dashboard, comparing key details between highlight and non-highlight works to see if a correlation can be established. The general design of the dashboard follows a format of comparing the highlight (red) and non-highlight (grey) works by the different field comparisons (e.g. country of origin), then a description describing the key insights to guide the reader.
Insights
Unfortunately, little can be drawn in terms of definite conclusions in this dashboard’s current state. Because of the limitations within the country and artist name fields, not many conclusions can be seen. The only trend that appears to be evident is that highlight works don’t take long to make compared to non-highlight works. To allow for definite conclusions to be drawn, a deeper analysis would be needed into other aspects of the works, or data would need to be supplemented for the missing fields.
Conclusion
I found this dataset quite interesting, but it wasn’t the most easy to work with since the data itself contain a considerable amount of NULL values for a good number of fields. This just served as a limitation, though, as I was still able to create a dashboard for my topic, even if the results were inconclusive.
Thanks for reading, if you would like to check out my dashboard feel free to check it out here.
