Inside Scoops from the 2017 Wolfram Technology Conference
November 1, 2017 — Swede White, Media & Communications Specialist
Two weeks ago at the Wolfram Technology Conference, a diverse lineup of hands-on training, workshops, talks and networking events were impressively orchestrated over the course of four days, culminating in a one-of-a-kind annual experience for users and enthusiasts of Wolfram technologies. It was a unique experience where researchers and professionals interacted directly with those who build each component of the Wolfram technology stack—Mathematica, Wolfram|Alpha, the Wolfram Language, Wolfram SystemModeler, Wolfram Enterprise Private Cloud and everything in between.
Users from across disciplines, industries and fields also interacted with one another to share how they use Wolfram technologies to successfully innovate at their institutions and organizations. It was not uncommon for software engineers or physicists to glean new tricks and tools from a social scientist or English teacher—or vice versa—a testament to the diversity and wide range of cutting-edge uses Wolfram technologies provide.
A Brief Data Analysis of the Conference
Attendees traveled from 18 countries for the experience, representing fields from mathematical physics to education and curriculum development.
One hundred thirty-nine talks were divided into five broad tracks: Information and Data Science; Education; Cloud and Software Development; Visualization and Image Processing; and Science, Math and Engineering.
We can take a look at talk abstracts by track using WordCloud. See if you can guess which ones they correspond to.
If you guessed Data Science, Science/Math/Engineering, Cloud/Software Development, Education, and Visualization/Image Processing from left to right by row, you have a keen eye.
We can also look at all talk abstracts divided into nouns, verbs, adjectives and adverbs. Perhaps a Wolfram Technology Conference abstract generator could be built upon this.
It was most impressive to see those at the conference who use the Wolfram Language for data science, education or medical image processing be able to ask questions directly to the R&D experts and software developers who make those tools possible for them. As Stephen Wolfram said, “It’s fun to build these things, but it’s perhaps even more fun to see the cool ways people use it.”
A highlight of the conference was the keynote dinner in honor of the 2017 Wolfram Innovator Award winners. Nine individuals and organizations from finance, education, oil and gas, applied research, academia and engineering were represented. We’ll have more on these outstanding individuals in a forthcoming blog post. For now, a tease of where they came from.
The conference kicked off with hands-on training, where attendees received individualized instruction on how to use the Wolfram Language for their research and professional projects, led by experts in Wolfram technologies for data science and statistical analysis.
Abrita Chakravarty, team lead of the Wolfram Technology Group, guided participants through a deep dive into data science with a morning session focused on project workflows, followed by an afternoon workshop devoted to data wrangling and analytical methods. Among the Wolfram Language functions highlighted were FeatureExtraction, DimensionReduce, Classify, Predict and more of Wolfram’s sophisticated machine learning algorithms for highly automated data science applications.
Take a look at Etienne Bernard’s (lead architect in Wolfram’s Advanced Research Group) recent blog post “Building the Automated Data Scientist: The New Classify and Predict” for further explanations of new machine learning features released in Wolfram Language Version 11.2. You can also view a livestream of Etienne demonstrating these features on Twitch.
In addition to hands-on training in data science, Tuseeta Banerjee, a Wolfram certified instructor, led a morning workshop on applied statistical analysis with the Wolfram Language. From hypothesis testing using DistributionFitTest to automated modeling using GeneralizedLinearModelFit, among many other functions, attendees were given the tools necessary to tell a complete analytical story from exploratory analysis and descriptive statistics to predictive analytics and visualization.
Wolfram U has on-demand courses available in data science, statistics, programming and other domains if you’re interested in learning how to use cutting-edge methods and the largest collection of built-in algorithms available in the Wolfram tech stack.
Stephen Wolfram’s Keynote Address
A highlight of the conference was Stephen Wolfram’s annual keynote talk, which covered an incredible amount of information over two and a half hours of live demonstrations in the Wolfram Language.
Celebrating 30 years of R&D at Wolfram Research since the company was founded in 1987, Stephen noted that’s about half the time since modern computer languages were invented. Next year, Wolfram celebrates 30 years of Mathematica—it’s fairly rare for software to remain so widely used, but the sheer amount of innovation that has gone into the product ensures its longevity.
Stephen highlighted some of the many new features in Wolfram Language Version 11.2 and noted ImageIdentify, which was announced in 2015, is at once a pioneer in general-purpose image identification neural networks and still paving the way as a building block in new Wolfram technologies. The neural net has been trained so well it can identify a jack o’ lantern carved in the fashion of Stephen Wolfram’s Wolfram|Alpha person curve.
From there, Stephen touched on everything from cloud and mobile apps to blockchain, with some examples of their uses for individuals, organizations and enterprise. It was a fast-paced, quickfire presentation that covered two Wolfram Language version releases (11.1 and 11.2), the Wolfram Data Repository, SystemModeler 5, Wolfram Player for iOS, Wolfram|Alpha Enterprise and a slew of upcoming functionalities on the horizon. He also hinted at how some large and well-known companies are using the Wolfram tech stack in ways that allow millions of people to interact with them on a daily basis.
Stephen told the audience about ongoing efforts in K–12 education and Wolfram Programming Lab, along with resources available to individuals of all ages and organizations of all sizes to learn more about how to use the Wolfram tech stack in their work and projects.
He also pointed to his recent livestreams, open and accessible to anyone, of Wolfram Language design review meetings—a rare glimpse into how software is actually made. You can view the collection of on-demand videos here.
Wolfram R&D Expert Panel
Each year at the Tech Conference, a panel of Wolfram R&D experts preview what’s new in Wolfram technologies and what’s to come. This gives attendees the unique opportunity to learn how Wolfram technologies are made, along with the ability to ask the people who pave the way of innovation at Wolfram questions about future functionalities. This year’s panel included:
- Arnoud Buzing, Director of Quality and Release Management
- John Fultz, Director of User Interface Technology
- Roger Germundsson, Director of Research and Development
- Tom Wickham-Jones, Director of Kernel Technology
Roger gave an overview of the hundreds of new functions introduced in Wolfram Language 11.2, along with hundreds more improved functions that are continually in development.
Perhaps one of the biggest highlights was a preview of the Wolfram Neural Net Repository, which provides a uniform system for storing neural net models in an immediately computable form.
Including models from the latest research papers, as well as ones trained or created at Wolfram Research, the Wolfram Neural Net Repository is built to be a global resource for neural net models. Classification, image processing, feature extraction and regression are just a few clicks away using the Wolfram tech stack.
Let’s look at some of the more creative uses of the Wolfram Language presented at talks during the conference.
Creative Highlight Number 1: Marathon Viewer
Jeff Bryant and Eila Stiegler demonstrated how the Wolfram Language can be used to analyze races and marathons, using the Illinois Marathon as an example.
Using data from the race, they showed how using functions like Interpreter can make the pain of wrangling and cleaning data easier. Jeff and Eila were able to take the data and create an animation that shows each runner’s progress through the marathon, with dynamics that indicate volumes of runners at any given time and location. Not only is this incredibly useful for people virtually tracking the progress of runners, but it also has applications for city and urban planning.
Creative Highlight 2: Building an Interactive Game Modeled on Jeopardy!
Robert Nachbar, project director with Wolfram Solutions, demonstrated an interactive game of Jeopardy! built in the Wolfram Language that he modestly said took him about a weekend to build.
Attendees were impressed with its functionality and excited about its application in education. Using built-in Wolfram Language functions like Dynamic and interactive buttons, Robert showed how an API call can be used to create a game of Jeopardy! with existing clues or how a custom game can be built. To demonstrate, he used clues and questions specific to Wolfram Language documentation. Toward the end of the presentation, a brief game was played providing clues to Wolfram Language functions that the audience could then respond to, providing a nice model for learning any topic one might think to program into the game.
Creative Highlight 3: Food Data in Wolfram|Alpha
Andrew Steinacher, developer in Wolfram|Alpha scientific content, gave his third talk in as many years on food and nutrition data in Wolfram|Alpha. The Wolfram Language now has nearly 150,000 searchable (and computable) foods. New computable features include PLU codes, used by grocery stores worldwide, and acidity levels, full nutritional information, ingredients and substitutions, along with barcode recognition for better alignment with international foods.
Future goals for food data in the Wolfram Language include better food and nutrition coverage for the rest of the world, specifically Asia; adding more packaged foods and more available data, such as storage temperatures, packaging dimensions and materials; aligning ingredient entities to chemical entities; new FDA nutrition labels with support for multiple sizes/styles; and computational recipes, including food quantities and nutrition, actions, equipment and substitutions. One can easily imagine how these tools will certainly innovate the food production and food service industries.
Creative Highlight 4: Presenting Presenter Tools
In a something of a meta-talk, Andre Kuzniarek, director of Document and Media Systems, gave a presentation of Presenter Tools, an upcoming feature of Wolfram desktop products. In his talk, he showed how talks created in Wolfram Notebooks can be prepared and presented with convenient formatting tools and dynamic content scaling to match any screen resolution. While some of this functionality already exists in the Wolfram Language, this improved framework elevates presentations to a new level of aesthetics and interactivity.
Wolfram Livecoding Championship
A fun evening highlight of the conference was the Wolfram Livecoding Championship led by Stephen Wolfram. For the contest, Stephen gave challenges to the participants, and they were then tasked with finding a solution to the problem using an elegant piece of Wolfram Language code.
Approximately 20 participants took part in the contest and responded to challenges ranging from finding digit sequences in π to string manipulation to finding the earliest 2016 sunrise in Champaign, Illinois.
Jon McLoone, director of Technical Communication and Strategy at Wolfram Research Europe, took home the prize for the most solved challenges.
The event was streamed live from Wolfram Research and Stephen Wolfram’s Twitch channels, and you can watch the video-on-demand here.
Wolfram Language Logo ImageRestyle Competition
This year, a new contest was introduced to see who could use ImageRestyle, new in Wolfram Language Version 11.2, to generate the most creative and interesting version of the Wolfram Language logo. Over 70 entries were received, and contestants were required to use the logo and another image or images to come up with a new machine-generated version of “Wolfie.”
This year’s winner was Emmanuel Garces Madina for the following submission.
More Wolfram Technology Conference Posts to Come
Chris Carlson, senior user interface developer, will present a recap of this year’s One-Liner Competition. Also, technical writer Jesse Dohmann will introduce this year’s winners of the Wolfram Technology Innovator Awards.