The limit, the outer edge…that queasy feeling in your stomach. It’s always the same, that moment of fear and unknown just before you jump and ultimately grow.
In social and data in general, analysts, scientists, and researchers are looking for that new edge, that new insight, that new finding in hopes of blowing something up to say they did it. As the framework that feeds this data, we often find ourselves in the amazing intersection of innovation and failure, and we want to keep falling. Ultimately this is driven by our customers as we are simply the exploration, collection, and distribution framework to help our clients push that limit forward.
Today we’ve decided to open up our proverbial books and share some of the things that are keeping us up at night and the problems we are trying to solve.
We’ll be sharing a series of interviews over the next few weeks by our team members, Brandwatch, Beanstalk Predictive, Zignal Labs, Social Intelligence Lab, Blackbird.AI & The SocialStudies Group.
To get things kicked off, we’ve interviewed Kaylin Linke who’s experience in social data is vast, from her time at early social analytics leader Collective Intellect to being a fundamental change-maker at Oracle and most recently building our own sales engineering and consulting practice here at Socialgist. Her goal and focus is to ensure our clients and the industry as a whole get the most out of our platform.
Hi Kaylin!
Thanks for agreeing to the interview and being the driving force for our customers to drive innovation and pushing their limits to get the most out of their platforms for their customers.
What is the bleeding edge for with regards to data?
Moving beyond just an industry buzzword or high-level concept, we are starting to see algorithmic fairness become a component of any data project. Fueled in part by highly publicized examples in the criminal justice system, hiring practices, and a rogue racist bot, this field of study is rapidly evolving with differing perspectives on how to tackle inherent and emerging biases. Personally I am encouraged to see the shift from identified issues toward practical approaches for addressing.
When it comes to the open web conversational data in our practice, I am seeing this shift illustrated by a greater focus and level of attention given to understanding the individual sources of data. Conversation and word choice are very different on a message board focused on motherhood and parenting compared to a gaming discussion board compared to a breakaway platform promoting hate speech.
What new dataset intrigues you the most?
More than one single dataset, my current favorite area of data exploration is the vast amount of conversational data coming from Asian markets, specifically China with sites like Weibo, Baidu, and QQ. As a non-Mandarin speaking westerner visiting a Chinese social media site is a sometimes overwhelming sensory experience but the view into daily life and learning layered meanings in communication patterns is well worth the challenge. I have had the joy of learning the phrase “sour lemon goblin” as a way of expressing jealousy in a self-deprecating way.
What findings could you derive from it?
The bridge from anecdotal insights to measurable analysis is an important one when it comes to data from this region. For those looking to cross, having expertise in language and culture is a must-have. In the lemon goblin example above without the cultural understanding of this trend, the proliferation of memes of lemons with faces might seem out of place. Advancements in the sophistication and accuracy of natural language processing are getting us closer but it is the combination of human + machine making these datasets a valuable resource for anyone conducting business on a global scale.
Thank you again, Kaylin, for your time!
Our next post will be from Paul Siegel, the Principal Data Scientist from Brandwatch.
