This past Sunday October 23rd we held our “1st Annual” Evolving Education with Cognitive and Data Sciences workshop before World of Watson. We knew this was a hot topic, but kicking off an event at 9am on a Sunday in Las Vegas had us questioning our sanity. We’re competing with football on a Sunday morning. Would anyone come?
The answer was crystal clear when we welcomed our keynote speaker Dr. Michael Karasick to an audience of 400 eager to learn about the Future of AI.
The Watson Analytics workshop featured Dennis Buttera from the IBM Ottawa lab and faculty from Australia (William Yeoh, Deakin U.), Canada (Ayse Bener, Ryerson U.), US (Haluk Demirkan, U. Washington-Tacoma, and the UK (Richard Self, U. of Derby). The panelists provided the audience with a wealth of insight regarding how Watson Analytics can be utilized in programs to empower every student to have the skills for smarter data-driven decision making. The discussion featured attendees asking questions about why Watson Analytics was selected and the outcomes developed by students. Each school had their own philosophy for deployment and outcome of skills developed that students can take into the workforce.
The Watson Cognitive API workshop featured Professor Gordon Pipa, Chair of Neuroinformatics Department at Universität Osnabrück who discussed his experiences integrating Watson into his courses and cognitively tracking and prediction flu progression. IBM Technical Mentor, Armen Pischdotchian gave a presentation and demo on the growing library of cognitive Watson APIs in Bluemix.
Over lunch we introduced the 3 finalist teams competing in the Watson Analytics Global Competition. Members of the teams flew in from Australia, Canada, and Slovenia. During the morning they presented their findings to judges who selected the team from Deakin University as the winner for their “air quality: towards a better future” project. Congratulations to Robyn Abel, Nakul Bajaj, and Scott Burden!
Following lunch we broke into 3 rooms. A workshop on the new data science experience (DSX), a workshop on Bluemix featuring the Internet of Things (IoT), and 5 invited research talks from academia and IBM Research.
The Making Data Science a Team Sport workshop led off with IBMer Armand Ruiz Gabernet who gave a thorough introduction to DSX. Faculty left very impressed with plans to immediately lever DSX in the classroom. Horst Samulowitz from IBM Research followed Armand with an amazing talk on his work building the first Cognitive Assistant for Data Science. The workshop closed with a brief talk on text analytics and SystemT by Huaiyu Zhu of IBM Research.
The Bluemix Internet of Things workshop featured Gayathri Magie from IBM, Dr. Gustavo Rodriguez-Rivera from Purdue University and Dr. Randy Hlavac from Northwestern University. The workshop provided an overview and demo of the Bluemix, IBM Cloud Platform as a Service and the Watson Internet of Things platform. The demo included creating an IoT application, visualization the connected device sensor data through dashboards, creating Real-Time-Insights on the sensor data and augmenting the IoT application with Cognitive computing capabilities by incorporating Watson API and include Weather data. Dr. Rodriguez-Rivera presented a novel way of looking at the web browser as the modern operating system powering applications, with Bluemix PaaS providing kernel-like services, and discussed how he is using this new perspective in the Systems Programming course at Purdue. Dr. Hlavac discussed how he blends IBM resources to create a unique learning experience for graduate and undergraduate on-campus and online students at Northwestern.
The invited research talks had a large and very interactive audience. Presentors included Aya Soffer from IBM Research, Winslow Burleson of New York University, Eleni Pratsini from IBM Research, Pekka Neittaanmaki from University of Jyväskylä and Jane Russenberger from Guiding Eyes for the Blind. The talks encompassed a broad spread covering how cognitive computing and analytics can be applied in a variety of novel ways with both humans and guide dogs. Aya considered how computational tools allow much broader access to data with some risk of hiding biases. Winslow presented an exciting approach to experiential learning through virtual reality experiences beyond current expectations. Eleni gave us new insights into cognitive analytics as applied to data captured through the growing Internet of Things. Pekka discussed how the University of Jyväskylä has integrated cognitive computing tools and techniques into their educational programs. We closed with Jane showing us how cognitive tools help identify which puppies had the greatest chance to be trained into being a successful guide dog. Each speaker answered questions until time had to be called.
During breaks attendees could meet with the Watson Analytics competition finalists, with students from the Illinois Institute of Technology who are building a website to explore the data & analytics job market using data from Burning Glass, with representatives from Xprize to learn about the Watson AI challenge, with reps from Kivuto and IBM to learn about the new academic initiative, and the team from Big Data University to learn how to lever data science courseware for academic and professional education.
We then moved on to 2 parallel sessions focused on data science education and cognitive science education.
The data science education session featured 5 speakers who covered a variety of important perspectives:
Cat Nikolovski of Hack Oregon kicked the session off covering her experiences helping volunteers productively work with open data solving social good problems in Oregon. She quickly discovered volunteers have the passion, but by and large lack the skills to build data products. Which led to the creation of Hack University whose sole mission is closing the data literacy gap and building the skills needed to work in teams and solve hard problems.
Next up was Brian Fitzgerald from the Business-Higher Education Forum (BHEF) who shared a plan for releasing national research and publications guiding business and higher education as they wrap their heads around the need to respond to changing workforce needs in data science, particularly around the ‘enabled’ graduate. Foundational to this effort is their data and analytics competencies work with the group. This project is unique is its focus on the data science competencies needed by any student in any field of study. Brian also laid out a process for how business and higher education can seamlessly work together to build new pathways in data science. Brian was followed by Will Markow of Burning Glass to present an early look at data & analytics job market analysis sponsored by BHEF and IBM. Stay tuned for the release of the full report.
Yuri Demchenko from the University of Amsterdam took the stage to present the work of the Education for Data Intensive Science to Open New science frontiers (EDISON) project funded by the European Union. We were very interested and excited to learn that the EDISON team and the NYC Data Science Working group crafted a very similar set of competencies. We left the event with the action to work on unifying the projects.
We closed the session with a fantastic talk by Peter Fox of Rensselaer Polytechnic Institute (RPI) who shared the amazing progress at RPI to introduce data science to EVERY student at RPI. They piloted the program this past summer with the initiative launching in Fall of 2017. While most schools are launching a data science program of some type they typically are designed for very small cohort of 20-30 students, RPIs approach touches every student. I had to write that twice. Here’s a third time. EVERY STUDENT.
The Cognitive Education Invited talks were engaging and highly informative. Our outstanding subject matter experts started off with Dr. Jim Spohrer, IBM Director of Understanding Cognitive Systems discussing the Cognitive Curriculum of the Future. Following Jim was Dr. Soloman Darwin, Executive Director of the Garwood Center for Corporate Innovation at UC Berkeley who talked about empowering the bottom of the pyramid through co-innovation strategies. Dr. Nitesh Chawla, Professor of Computer Science and Engineering at Notre Dame spoke of curriculum evolution being driven by the nexus of society, data and cognitive sciences. Our final speaker, Dr. Artemesia Jaramillo, came all the way from Saudi Arabia where she teaches as a part the remote Dublin City University campus. Dr. Jaramillo along with IBMer Susan Malaika gave a compelling picture or their work to bring cognitive skills to the largest number of women in any STEM program in the world.
We closed the day with a panel session featuring Guru Banavar from IBM research as moderator, Girish Punj from University of Connecticut, Brian Fitzgerald from the BHEF, and Gordon Pipa of Universität Osnabrück. Guru explored multiple topics centered on the theme of our day long workshop. How should education evolve with cognitive & data science. We recorded the panel and hope to have it posted to YouTube before US Thanksgiving.