Skip to end of metadata
Go to start of metadata

Please join Research IT and partners as we discuss visualization technologies and services on campus, and the community that will need to form in order to have a significant impact in this space on research, teaching, and learning. During this conversation, we will survey the landscape of visualization on campus and in higher education, and we will develop an initial framework for describing the range of applications that should be included in the framework. This will include a broad examination of what visualization means in different disciplines, as well as a discussion of opportunities across traditional boundaries. The group discussion will also cover a variety of needs that are emerging, from software licensing to computational environments, consulting, and training.

When: Thursday, 30 November from 12 - 1pm
Where: 200C Warren Hall, 2195 Hearst St (see building access instructions on parent page).
What: Exploring visualization services for UC Berkeley
Presenting: Chris Hoffman, Research IT; et al.

Prior to the meeting, please review:

Presenting: Chris Hoffman, Research IT


Aaron Culich, Research IT
Amy Neeser, RDM (Research IT & Library)
Andrew Wiedlia, LBNL
Anna Sackman, Library
Anthony Suen, BIDS
Barbara Gilson, SAIT (emeritus)
Becky Miller, Library
Chan, Library
Desi, Library
Elliot Smith, Library
Indu Tandon, OHR
Jean Fergusun, Library
Jenn Stringer, Research, Teaching, and Learning
Kali Armitage, IST - Document Management
Kortney Rupp, Library (Chemistry)
Larry Conrad, CIO
Marlita Kahn, TPO
Mark Chiang, IST-EDW
Owen McGrath, ETS
Patrick Schmitz, Research IT
Quinn Dombrowski, Research IT
Rick Jaffe, Research IT
Steve Masover, Research IT
Yasmin AlNoamany, Library


Slides (PDF)

Not a presentation -- a conversation about how we can bring a visualization service into being!

Q: Why community discussion to build service?
Elliot: focus on what's needed, not duplicating effort
Larry: differences between disciplines, can't get total picture unless looking through lens of differnt disciplines

Q: Is viz different in Data Science?
Rick: Public Health -- connecting data to action. Differnent from exploring the data more generally. Purpose of viz influences.
Patrick: Communicative vs. Understanding -- different goals & audiences, related to what Rick just said.
Jenn: Where does HearstCAVE fit in relation to scientific viz?

Q: Is viz mainly about building effective figures?

Q: Viz via Matlab vs. Tableau?
Anthony: Open source solutions is where lots of data science viz is heading
Patrick: If open source produced better results, why would vendor packages exist?
Yasmin: Vendor packages tend to be easier to learn/use. Students I've worked with like Tableau, struggle with R.
?: Lower barrier to entry
Larry: super-esoteric, finely tuned expertise; vs. tool with broader application (and, often, access)
Patrick: Think of desktop publishing "revolution" -- access to tools didn't make everyone a good designer ... how you use color, how you use dimensionality -- just on communication side -- there's a real skill involved, and tools don't automagically confer skills.
Aaron: MatPlotLib - new color default palette to help people avoid selection of a palate that communicates poorly (link)
Amy: My experience also, when I was a plant bio librarian
Chris: Viz literacy issue
Patrick: pure viz designers who are embedded with scientists help communicate better AND have a positive impact on the scientists who work with them
Jenn: Skills/Roles -- a viz designer who specializes in science/data communication
Yasmin: Some of what the tools offer are poor choices for communicating information

Q: Reading - scientists need data viz

Q: Is information viz different from data viz?
Patrick: Sorry Marti (School of Information) couldn't be here. Worth thinking about this difference: info viz is for communicating something you know (have discovered); data viz is about discovering what information is latent in a data set. But Marti would have more to say about this.
Anthony: Best viz components allow display of detailed, curated information, and to see the data in various overviews.

Q: What's happening in VR on campus?
Jean: Maker space in Moffit bought some rigs and do open-houses giving students a chance to try them out.
Jenn: Video production unit purchased some 3D viz tools
Steve: I saw a demo a couple CalDays ago at Jacobs
Aaron: An area where there's interesting exploration going on: using augmented reality in association with emotional intelligence research -- have seen some folks coming to campus to talk about this, but haven't seen activity in this space at UCB yet
(?): VR in MEng
Anthony: Data viz with VR -- not ready for prime time yet -- on us to figure out what is effective, what is best to teach and experiment with in this space.
Patrick: distinguish between research about VR vs. research that employs it in pursuit of some (non-VR) research goal
Patrick: new medium for storytelling requires training in how to communicate well in that medium
Larry: UNC Chapel Hill -- predecessor there built new IT building, built 3 rooms for viz ... weren't used ... no one knew what to do with them
Marlita: On the learning side, is there a distinction between teaching how to use tools (create viz); vs. teaching how to use them in a particular topic or to a particular end.
Marlita: Big History -- a tool seen here a few years ago -- there was talk about building an augmented-reality component to the tool, not sure where that went.

Q: What about geospatial viz? Do we separate GIS when we consider viz, given how much is going on in that specific space?

Q: What else is happening on campus?
* (DLab is hosting this). Includes a landscape scan (see list of orgs on the site).
* Patrick: Dani I @ LBNL / EECS -- ImageXD at BIDS -- does a lot of work around viz
* Patrick: LBNL has a group of experts who help scientists with viz
* Larry: sense of researchers here on campus interested in viz? Patrick/Quinn: we've worked with astrophysicists making fabulous vizualizations using Savio; materials science, chemistry, climate science -- all areas where viz is important.
* Anthony: data science pillars: machine learning, inference, viz. But there are no real standards, many tools, a nice to have at the end, no center -- all this presents structural challenges to how to teach/support
* Patrick: administrative use of visualization -- Tableau Users Group is grounded in this user base
* Kortney: have faculty who have interest in viz; also: students don't know to ask a Chem Librarian for help in viz space
* Chris: part of point of site / email is to create a central place people can ask questions and be productively referred
* Aaron: UC Extension - Data analytics course, resource we can use. Also: Berkeley Science Review -- this is a place where people (students) are trying to do viz and could be connected with resources.
* Jenn: EDW, ETS -- the latter, learning data, maps, creating connections among and for students vis-a-vis learning experience
* Patrick: Data Science fairs early on surfaced a surprising breadth of DS going on on the campus, wonder whether the time is ripe for a Viz Fair
* Larry: Where could the campus make an investment in this space that would be productive
* Indu: user interface design need, to points made earlier -- teaching people the skills with which to use the tools
* Jenn: a place where people know to go, including support mechanisms for teaching skills. An institution that does this well?
* Patrick: U Michigan
* Marlita: GIGO - valid use of viz to accurately describe a data set
* Patrick: This question of validity and accuracy is a central aspect of Data Science curriculum.
* Anthony: An ethics component to Data Science major program
* Larry: Hearing that design, skills training is needed?
* Patrick: bringing the resources together, providing the expertise more broadly
* Amy: Michigan has a data viz librarian -- a position just for that -- constantly booked, high demand on his time.
* Patrick: having a physical space a UM is
* Kortney: what is meant by viz is often. diversely or poorly understood ... maybe a fair with an aspect that awards best design, best use of tools to communicate a dataset/result.
* Marlita: Erudite - mining an organization's information resources to surface expertise where it's not necessarily visible across the org. They're looking for use cases, perhaps surfacing viz @ Berkeley knowledge (or other types of expertise) might be of interest
* Becky: Duke University is another university that does Viz well
* Jean: NC State
* Jean: consider what can be done with the bottom three floors of Moffitt Library ......

Aaron: AEoD as a platform for offering tools to many while purchasing only a small number of licenses
Chan: UCOP has bought Tableau license/server -- worth looking into this.
Mark: UCSC -- interested in sharing license for Tableau

  • No labels