Behind the Scenes: Vizzing for GovTech’s DAT2022
This year, NLB’s Data Office took part in GovTech’s Data Arcade Tournament (DAT), a whole-of-government visualisation competition. We are excited to share that we have emerged as the winning team, competing against more than 500 teams! So here we are, sharing a bit of the behind-the-scenes on how our team built our dashboard entry.
The Vizzing Process
First things first, my teammate Guo Jia and I sat down to discuss what to work on, keeping in mind that this year’s theme is ‘Twist and Turn’.
One of our initial ideas was to analyse how borrowing patterns had changed pre- and post-covid. While that is impactful, we speculated that many other teams will probably work on covid-related data. Plus, in another blog post, we have already covered some aspects on how Covid-19 had transformed our patron’s borrowing behaviour.
3 coffees later, we decided to go ahead with something that resonates with us (and hopefully for you as well), which is to analyse data revolving our Singapore collection (literary works written by Singaporean writers or authors).
What goes behind the building of a dashboard?
While we are very proud of how our dashboard turned out, it definitely didn’t start out this way.
In fact, our first iteration looks almost nothing like what we have now, and probably shouldn’t see the light of day.
Next, we will test to see if the interactivity works / make sense, before working the aesthetics bit at a later stage. Continue reading to see how this turned out!
Deciding what we want to include in the final dashboard were based on a variety of reasons:
- Does the chart reveal anything interesting?
- Is the data easy to obtain / clean? In some cases, we might want to give up and search for alternative datasets instead.
- Is the final dashboard too cluttered? It is often easy to dump many charts into a dashboard, without considering its usefulness or contribution to the entire story.
- Does the storyline flow nicely from one tab to another? If not, what can we include to tie them in together?
When it comes to dashboard interactivity, it is important to include some text to guide users. As part of our entry, we had done up a SingLit recommender, where we recommend users SingLit titles according to what non-SingLit titles they will like. For eg, users who enjoyed Becoming by Michelle Obama, will be recommended titles such as This Is What Inequality Looks Like by Teo You Yenn, as it is also tagged under ‘Politics’.
It might not be intuitive to users that they should click on the book covers, so including texts to inform users to do so is definitely helpful.
(Did you recognise the previous iteration of this recommender? Yes, we painstakingly added the book cover images for each title 💀)
The cherry on top
To tie everything together, small details such as the colour palette, font, and visuals play a very huge part in making the dashboard cohesive. And although not very noticeable, we also designed some visuals to further tie back to the ‘Twist and Turn’ theme.
A final note
Dashboard creation is like an artform — there is no right or wrong. So along the way we did make further improvements to our dashboard! We have added screenshots of our final submission below. We welcome any feedback so that we can all learn from this together too!
Text by Phyllis Tay
National Library Board
