Continue reading »]]>

Hello, THATCampers! In the spirit of one session/one tool THATCamp production, this document is the outcome of a discussion about how to prep data and choose a visualization tool for humanists who don’t have direct access to high-level database and programming skills.

We welcome your comments, suggestions and feedback. Hopefully, you’ll find this step-by-step planning document useful as a hand-out to communicate with colleagues and/or students who are just getting started in the research process for projects that include data visualization. Additionally, there is a list of visualization tools collected in the Google Doc that this session’s attendees produced.

Thanks to all of the session attendees for lively discussion and great contributions!

Steps to take for managing data to make visualization easier:

1. Start with the argument you’re making and how that argument could look on paper
   1. Start thinking about the visualization before you start taking notes
   2. Find visualization tools to match your paper mockup
   3. Determine feasibility of collecting data for that visualization tool/type of visualization
      1. This will depend on your skill level and comfort with technology.
   4. Understand the limitations of the visualization tool (e.g. single date required when data is often in date ranges)
2. Consider end result
   1. Data exploration? Out of the box software is best used for exploring data
   2. Data presentation in argument form? Building an argument in graphic form will probably require (but check with @tjowens about facets in Recollection)
   3. There will be a $0, $1,000, $10,000, $50,000, $100,000 and $1,000,000 version of this. First make the one that costs nothing and think about how you would scale up if it turned out to be particularly interesting.
      1. If you compromise and go a low-cost version, don’t forget the idealized version you wanted in the first place.
3. Find the least complex tool you need for the job of data collection
   1. Excel is a useful tool, but data with a lot of repetition is ideally expressed in a relational database.
      1. Excel can auto-complete entries, but auto-complete can also create inaccurate data
4. Start with small data sets, and iterate often
   1. Simplify data w/ data dictionary
   2. Use visualization as data remediation

Continue reading »]]>

This session is happening! Sunday, 11:00 am – 12:15 pm, Room 92. If you’re interested, please (please please please) read and comment on the Googledocs draft session structure!!!

There are a number of ongoing projects that center around structuring, storing, sharing and visualizing data within the humanities, ranging from well-known tools such as Zotero to brand new tools stemming from recent grants that are still being prototyped. These efforts create lots of opportunities, and sharing data between these tools and initiatives benefits the whole DH community. However, designing for and implementing data structures that support this kind of sharing adds a different kind of complexity.

The question is, then, how do think about structuring, organizing, and sharing our data going forward so that our structures are both flexible enough to hook into when we build new tools but structured enough that the data sets would talk to each other? How do we tie together different kinds of data sets (for example, but not limited to: GIS, citation management, prosopography, timeline and event tracking, etc.) in a way that works across several disciplines? How do we structure the data so it integrates well with visualization tools? What are the benefits, costs, and challenges of an undertaking of this kind?

If we break it down even further, we can ask more granular questions about the data we collect when we do research. What kinds of data sets do you have? What kinds of data show up in those sets? What kinds of relationships do you want to analyze between those different kinds of data? How do these questions change (or stay the same) across disciplines?

While it’s not easy to answer questions of this scope in a single session, THATCamp’s unconference format seems like the ideal place to start!

Continue reading »]]>

We put stuff on maps. We can put stuff on historic maps. We can put things on timelines. What do we want to do next? There are a ton of cool visualization things going on, many of which use simple javascript libraries. Which of these are useful for cultural heritage collections? Further, what kinds of ways of visualizing are useful for what kinds of objects and collections?

In part, I think this is about a bit of a translation between the very active world of data visualization into a context where we think about various digital incarnations of cultural heritage objects. So, visualizations as interfaces for discovery of individual things as well as visualization in the more general sense of seeing the big picture. I would be interested in setting up a bit of mind meld to think through some of this. Specifically, what kinds of data do we have and what are some new ways we can create web based interfaces and visualizations to interact with and understand that data?

Here is a suggestion for how we could go about this:

1. (20 min) of swap and shareVery quickly look at a bunch of examples in these three categories: (For starters we can start listing things we want to talk about in the comments on this post)

• a) sites with cool visual interfaces to cultural heritage collections./li>
• b) examples of sets of items, individual objects, and other stuff that we want to interface with./li>
• c) fun visualization widgets and interfaces that have nothing to do with the humanities but that we might mine for ideas or ideally just use.

2. (30 min) Small groups sketch out one page wireframesI would then like to break into small groups and match make some of the As Bs and Cs. Take some of these cool ideas, pick a specific kind of collections and have each of the smaller groups create a one page small paper prototype or simple wireframe of what this cool interface would look like with a particular kind of digital collection or digital object.

3. (10 min) Rapid report outWe would then take 2 minutes each to report out the pitches for these interfaces. (Could video these on someone’s phone and post to youtube?) We then post pics and one paragraph descriptions of what this interface could do for a specific kind of collection or kind of object.

I have included some initial examples of stuff we might talk a bout in part one. If your interested in this please add more ideas in the comments.

A) Example interfaces:

• Mappy timey stuff like historypin: www.historypin.com/
• Historical data on maps: acl.arts.usyd.edu.au/harlem/
• Change over time in editions of a text: benfry.com/traces/
• Railroad owner thing: nexthistory.org/shaping-the-west.html
• BYU Corpus Stuff: corpus.byu.edu/

B) Fun Data set/item examples

• menus menus.nypl.org/
• pictures loc.gov/pictures
• quilts www.quiltindex.org/msumuseumcollection.php
• sheet music data.gov.au/dataset/south-australian-sheet-music-state-library-of-south-australia/
• convict transportation records data.gov.au/dataset/british-convict-transportation-registers/

C) Open Visualization Libs

• Timeplot: www.simile-widgets.org/timeplot/
• InfoViz thejit.org/demos/
• TufteGraph xaviershay.github.com/tufte-graph/
• milkchart code.google.com/p/milkchart/
• VIDI www.dataviz.org/

So what do you think? What things should we add to the list? Who want’s to be in on this?