One hundred trillion — that’s not the U.S. national debt; it’s how many data points Adobe tracked last year. That number is larger than the total number of humans to ever live. All to optimize customer experience and engagement.
Big Data’s Big Challenges
When people talk about big data, they often don’t realize just how big that data really is. And with a vast amount of data comes big challenges. Anil Kamath, an Adobe fellow and vice president of technology, explains, “With all that data, you run into the challenges of large volume, large variety, and high velocity.” It’s a problem common to today’s businesses, and Adobe Experience Cloud is among those working on solutions.
We sponsor research to help solve some of the challenges of big data. On June 1, 2017, Adobe kicked off its fourth annual Data Science Symposium. The Symposium is the annual capstone for a series of grants and internships that promote an industry-academia partnership and serves to highlight research into technologies like machine learning and AI — technologies that are solving the difficulties of big data and helping brands provide customers with the best possible experience.
Adobe’s 4th Annual Data Science Symposium
Part of the challenge of understanding big data is that it’s like trying to build a water-filtration system under a waterfall. No matter how hard and fast you work, the data just keeps pouring down at a crushing rate, overwhelming your efforts. Anil explains, “Marketers can’t keep up by manually sifting through and analyzing data. You need machine learning to get insights. You need data science to extract signals and make predictions about the best marketing efforts — how to deliver the right content to the right person at the right time.”
Companies and researchers across the globe are working to address the issues, but, traditionally, most institutions work on their own. In academia, you have to overcome tunnel vision and think about real-world application. In business, companies guard their trade secrets to maintain a competitive advantage. Edward McFowland III, Assistant Professor of Information Systems and Decision Sciences at the Carlson School of Management, University of Minnesota, admits, “We spend a lot of time in our offices, working through questions and problems, and sometimes we end up having a very micro-view of the world because we have a very specific question.”
However, Professor McFowland is bucking the trend by partnering with Adobe as he studies anomaly detection, which he explains as “identifying systematic patterns that are indicative of a change or some kind of event of interest.”
Adobe’s Data Science Symposium is designed to give researchers like Professor McFowland an opportunity to share what they’re doing and gain insights from others working on similar topics. Anil says, “The Symposium is aimed at sharing with academics who are working in the field of data science. We want to inform them about the area of digital marketing and the kind of problems that digital marketing has that are relevant to data scientists, machine learning, and AI.”
This makes the Symposium important for more than simply being a forum to present new research. The Symposium represents a partnership between industry and academia — two worlds that, traditionally, have been apart.
The Next Evolution of Research
Hema Yoganarasimhan, professor at the University of Washington, attended the Symposium and is studying the shift in software from the traditional, perpetual-license model to the software-as-a-service (SaaS) model. Her research will help companies understand customer turnover and usage patterns, and she recently secured a $50,000 grant from Adobe to help fund her research. Even more important than the grant, she and her students also get access to Adobe’s vast collection of data — data critical to performing research and developing new innovations.
Speaking about the benefits of this partnership, Professor Yoganarasimhan says, “First, we have access to all the incredible Adobe data and guiding questions. That really helps, and we can publish based off of that data eventually. Second, my students can get really good corporate experience that they would not get just being in the PhD program. That’s important. It’s a wonderful chance for them to get out and see the real world.”
For Adobe, academic researchers are a great source of new innovations and technology — if that research can be applied to real-world problems. Anil says, “We recognize that a lot of innovation in data science is in academia, and we want to help these researchers by making them more aware of the real-world problems in digital marketing. It’s also a way to get into the academic research to see how it can be useful to us.”
That research is used to guide the evolution of technology like Adobe Sensei, Adobe’s AI and machine learning framework, and making them smarter and faster. Anil says, “Our goal in tracking this data is to use it to enable marketers to design better, more relevant experiences.”
Collaboration is the End Game
Anil is a firm believer that collaboration is the key. “Innovation happens in industries and academia,” he says. “It’s about collaboration. Sharing real-world problems and applications, giving them real-world data, and having their students come intern with us to get real-world experience. Doing those things helps academics understand the kind of things industry needs to solve. Additionally, the industry benefits by leveraging academic innovation and the student pool to work on these problems. So, that is the major emphasis of the symposium: Sharing.”