|Shruti Shetty - Art Director & Data Science enthusiast|
Mauricio: Why your passion for Design and Data Science?
Shruti: Design has gone through quite a silent transformation over the years. The canvas, the technology, the aesthetics have all evolved significantly. Now more so than ever I think, design plays a very central role in tying all these changing features and packaging them for experience. Moving into the new age of the Data Economy, made me look and explore this new and next avenue that will impact design in the coming few years. This lead me to Data Science.
M: Data Science brings insights from analysing data sets. Design can then help present those insights on ways they are easily understood but also engaging and interactive. As a result, those new insights or findings from data sets can highly resonate with audiences...is that correct?
S: That's correct. Traditionally data analysis and handling has taken place in the development or mathematical sense of the world. With the help of design, we translate those complex models, from texts and numbers to patterns, which we can see and understand, and in context to a collateral situation or problem. As said, an image might be as powerful as thousand words.
M : Analysis of data is just one part of a story; in some cases investigative journalism, historical context, current market, political, economic and social dynamics have also a great influence on helping shape data into powerful insights. How do you see the future of Data Science with other disciplines and how Design can play a key role?
S: Patterns are everywhere. They are the very fabric of human behaviour, which has now translated to our online lives. With the power of technology and our need to be switched-on 24/7, we are contributing continuously to this massive data pool. This also means data is being continuously altered and influenced by all and/or more factors, than the ones mentioned above. Hence, before we even begin processing the dataset, we try to understand the circumstances in which the data has been collected. Sometimes this can help fill in the blanks, with the bigger picture. The purpose of data science is to help tell our story, with the events & features of our data. As the scene setting changes so does the climax of the story.
Itcan be quite hard to identify these influencers in the process of crunching numbers, and this is where visualisation comes into play, within the process itself. Visual tools also help determine the influencing features and charter an appropriate course for the story.We use tools like graphs, charts, plots etc. that display the workings of the data sets in a visual form. As mentioned before, design is a key interpretation of this story into a visual language, which everyone can understand.
M: How important is for Designers (who help Data Scientists) to be involved from the early stages of data analysis? so they can fully understand final insights and create solutions that matches what Data Science want to convey...
S: Designing or visualizing isn't an end mean or a final output stage. Many designers would agree, that design and creation is a journey and a process by itself. Such as data, design goes through it's own transformations, gradually. Hence for a desired optimised solution, collaboration between the two fields, is key. The key is to compliment and co-relate.
M: Is it the role of Designers to just help Data Science better visualise data insights? or the Design role goes far beyond that?... Can Design also help companies better collect data or help companies to find better ways to engage people to contribute to data? or even more, can design engage audiences to analyse data sets?
S: I began the studying data science, my intention was to implement this within our service pool, for our clients. The data we collect can help impact the strategy with which we can constantly improve our products or technology, given the patterns that our users or consumers display. Which in turn, gives them a better experience.
This evolution needs to be refuelled by new and more precise data, which can again be facilitated by design. In short, we use design to translate the given said data, and help grow and alter the data set, to maybe evolve into a new pattern. According to me, this agile approach would best suit this relationship, between data and design.
M: There might be plenty of data visualisation tools these days. What tools have you worked with and found them great to use?
|The Method Case: The project uses digital practices and processes to blur the lines between photography, data visualization, textile design, and computer science.|
M: Organisations with no access to Designers or Data Scientists have a big challenge when analysing and presenting insights from data? Any tips for them?
S: Ideally, gathering of that knowledge and datasets would be the first step, moving into this direction. Maintaining records and collecting them in a consistent and efficient manner. I guess with any given field, you need to have access to expertise on that subject matter. If you don't have those skill sets in-house, you would seek those services elsewhere. Traditionally, companies have been gathering the knowledge for the datasets, but haven't been able to implement it's full potential, back into their businesses. For businesses that want to explore this area, there are plenty of options out there. There are independent Data Science organisations who can then collaborate and partner with other agencies, to help make use of their data. But at the same time, companies like ours - Adrenalin Media, are upscaling their skill sets to include this as part of their services, even for Small / Medium sized companies to big scale corporations.
M: Do you favour tools that automate the discoveries of regularities in data? meaning less work for data crunching and possibly just more work for finding compelling ways to present those discovered regularities?
S: I have learnt the hard way, there is no easy formula to help clean up datasets. Again, softwares like R and Python are quite versatile and can handle this process quite well. Personally I like the Google Refine programme, as it has some really nifty features such as a log of all the changes you have made to the dataset, which helps reverse back any mistakes you might have made.
M: As cloud services become more popular and affordable, companies now have the opportunity to collect massive amount of data. However the reality is most of that data gets untapped and/or un-used? Apart from using Data Scientists to help companies find better ways to achieve their key objectives/goals by effective analysis and presentation of Data turned into insights; is there anything else companies should be using their data for?
S: My answer would be similar to the above.
M: There is a lot of talk about how to use big data on organisations and companies but what about individuals, citizens using data for everyday life activities or local community based activities? Any examples or insights on this front?
S: Definitely! Why should all the big companies have all the fun? Sites like help visualize your personal data into a cluster of visual graphics, aggregating all your stories online, from LinkedIn, Facebook, Twitter etc. Companies and services alike, have realised the power of visualization and there are a lot of them out there empowering us to take charge of our own data footprints to chart and describe the journey, in a visual form. It also helps keep things fun and interesting for the users.
M: Powerful visualisation of data is key but what about user interaction with data. Any user data driven projects and at what point can interactivity be a nuance for users to fully engage with stories and insights?
S: Interaction is such a powerful engaging tool. Data science wouldn't exist without user interaction. It's the basis of the records that we use. At the same time, the benefits of our insights are then used to display that information for the user to 'interact' with, in terms of engaging with and modifying a visual art piece or using a said service or product. The User interaction is central to the validation of the predictions that we as Data Scientists make. And design can then present this case for the user to accept or reject. In either cases, that would be fed back into the cycle, for the whole process to flow again. It's such a powerful concept, if one thinks about it.
M: What are possibly the top three projects you believe have implemented great design for their data driven projects?
S: Design can be assessed aesthetically or functionally. A good design follows a balance of both. Amazon, Netflix are a few of the companies that continuously update their design based on the data and user patterns that they record. While they aren't aesthetically prominent, they are studied to be one of the most effective use of Information Architectures. Similarly, you see startups such as Triptease, Evr.st etc. who hire data scientists, terming them as 'Growth Hackers' to bust and crunch the numbers and deliver visually stunning outputs for their users. The recurring themes with all these projects is the agile collaborative way that the data and the design teams work together to keep their services and their presence fresh and valuable.
|Triptease - For finding lust worthy destinations curated by users|
Shruti has a design background and been working in the digital space for the past 8 years now. After graduating from BillyBlue College in Sydney, with a degree in Communication Design, over the years she has specialised in User Experience and Information Architecture. In her current role as Art Director at Adrenalin Media in Sydney, her team produce online solutions, ranging from websites, campaigns to integrating new technologies into their services for eg: RFID.
|Infographic explaining Big Data, made by James West and published by New Scientist. It starts with the most fundamental concept which drives big data 3 Vs of Big Data and then moves on to talk about trends, features and challenges of Big Data.|