I recently attended a MeetUp of the Los Angeles Semantic Web & PHP Meetup Groups. It was a presentation on the FactualAPI. Factual is a very cool company building a platform for “open source data.” I pulled down some data to test and created the following visualizations, which reveal…well, something about American Presidents. I’m not quite sure what that something is, but it’s interesting to me. What do you think?
presidential_branches
cool_factual_President_years
presidents_in_USArmy
space
space
Interconnectivity of Presidents Over Time
This shows the shared connections to universities & branches of the military over time. I have grouped them into groups of 5 presidents at a time, which are the ones highlighted in yellow. This is the most interesting case I found. The middle of the 20th century appears to be the least interconnected.
factual_Pres-VP v4-4
factual_Pres-VP v4-3
factual_Pres-VP v4-11
factual_Pres-VP v4-10
factual_Pres-VP v4-9
factual_Pres-VP v4-5
factual_Pres-VP v4-6
factual_Pres-VP v4-7
factual_Pres-VP v4-8
space
space
Presidents & Their Colleges / Universities
This is interesting because it is not, as I would have hypothesized, all Harvard and Yale folks. In fact, the Harvard / Yale crowd is most best represented at the beginning of the country and in the last 50 years. A lot of presidents did not attend college.
Over the last month, I’ve been working with Edwin Bender and The National Institute on Money in State Politics (NIMST), which operates FollowTheMoney.org. This not-for-profit organization provides a database similar to the Sunlight Foundation’s database of political donations, but NIMST focuses exclusively on state elections.
This is an excellent use of this type of data. I encourage you to read the article and the report.
For a test run, NIMST gave me a sample set of data to process in Jute. Admittedly, I have not perfected the analytic techniques here and there is a lot more work to do to make the data relevant. Edwin Bender said to me in a prescient email
“Clay Johnson of Sunlight Foundation said recently that data visualizations are a dime a dozen, but meaningful visualizations are priceless.”
These are the top 5 donors to state candidates in FL in 2008. The largest donations went to the FL Republican Party. These donors give a large number of $500 donations (shown in orange).
space
space
So this is my first attempt at meaningful visualization. I give myself a “B-” on it. I took the data from FL, a state where I’m familiar with state politics after playing a prominent volunteering role with my mother’s unsuccessful run at the state house in 2008.* In the following visualizations, there is a very simple visual index: REPUBLICAN = RED; DEMOCRAT = BLUE; $500 DONATION = ORANGE.
The interesting about this is:
1) The top donors in the state are largely bi-partisan in their giving. Why? Do their recipients sit on influential committees? Is there a specific piece of legislation they wanted to support? A specific fundraising volunteer who brokered these donations?
2) There are a surprising number of $500 donations. Why so many?! What is the significance of a $500 donation to a state official? UPDATED: ”$500 is the legal limit for a campaign contribution to an individual candidate. That amount can be given once during the primary and once during the general election.” according to Linda McDonald, my mother.
Why? I’m not sure. I hope this blog post will help me find out.
FYI: you can access this data set in a network document called “state money FL v1″ by using the Jute guest account. Or, request your own account and I’ll share the data set with you.
Top 2 Donors in FL State Politics in 2008
These are the top 2 donors in FL state politics. What is their political bias?
space
Paul Tudor Jones, II
Mr. Jones supports candidates from both parties and both parties' state offices. Why?
space
Guy M. Spearman, III
Guy M. Spearman, III also supports candidates from both parties and both parties' state offices. Why?
space
Whole Network Analysis: Top 5 FL Donors
Whole Network Analysis shows a surprising number of $500 donations among the top 5 donors in FL state politics in 2008. $500 is the maximum donation permitted.
space
Assumed Industry Affiliation
By extrapolating industry affiliations from the companies where donors self-report that they work, we can look at trends of how industries' money flows to parties. This is not really a factually valid analysis--just an exercise in using Jute to crunch NIMST data.
space
Finding meaning in this data…
This data set is just a trial, and this blog post is just to get some initial feedback. My hope is that we can elevate this into a funded research project, analyzing trends in the NIMST data, and finding specific, relevant examples that we can publish. Also, I’d like to work with a campaign to help them in their fundraising and strategy development.
There is a lot more than needs to be done to “take this to the next level.” That includes:
Cross-referencing this data with other data sets, like: which campaigns actually won seats in the State Legislature; which committees the elected officials sit on and chair; which federal candidates these donors supported; and, ideally, which candidates I know (or my clients know) and how they can leverage existing relationships to get value out of this analysis. (That last one comes from existing, internal databases.)
But for now, I hope you’ll help me find meaning in this data. If you’d like access to more analysis from other states, or you are interested in publishing this material, please contact me via email or via phone at 828/545.9539.
Sean McDonald
co-founder, Jute Networks
space
*A little editorial here: Linda McDonald ran a hell of a campaign against a competitor with an exponentially larger war chest. In a district where a Dem hadn’t had a chance in a generation, she lost by a razor thin margin. And the people of the state of Florida are worse-off because she doesn’t represent them.
Independent Sector is the leadership forum for charities, foundations, and corporate giving programs committed to advancing the common good in America and around the world. Their nonpartisan coalition of approximately 600 organizations leads, strengthens and mobilizes the charitable community in order to fulfill their mission. Their constituents include some of the world’s leading non-profit organizations like the Gates Foundation and the Kellogg Foundation.
They rely on high-quality information about elected officials and their relationships to those officials. The first step–figuring out “who knows whom and how”–takes months and sometimes even years. With the help of JUTE, it takes just a few minutes to find a “pathway” from Independent Sector (IS) to key members of Congress. (Please note: the names have been changed, out of respect for the privacy of IS and their constituents.)
Jute found strong relationship pathways from Independent Sector to an Elected Official. This saves IS months of time in the “information discovery process.”
1st_path_highlighted_large
IS_first_degree
Example-_ IS Political Network v1-1
bridge_top_3_paths
partial_bridge
big_network
space
space
Details
IS became our client in Summer 2009. In the fall, our Nexus team rolled into Washington DC for an on-site project to help the IS executive team convert their slow, social process of discovering people in their networks who hold established relationships with elected officials into an efficient business process. This allows them to spend their time actually building relationships, not just figuring out what the most beneficial ones are.
This conversion, when complete, will lead to a massive improvement in efficiency for IS and help to uncover the wealth of “relationship capital” that the IS team, board members & member organizations maintain. After just one week, the results were impressive.
Elizabeth Kohm, Vice President of Resource Development, described the change as, “being able to go from 0-60 in five seconds, instead of five months.”
When we arrived on site, we had a few concrete goals:
Start with 6 data sets with over 200k records & mash them up into “database nirvana”*
Find “pathways” of established relationships from the IS team to a specific elected official
Generate reports that can be used by the executive team and board at IS
Get to work, accessing the information and requesting introductions along those relationship pathways
In the end, the IS project was a success. This was a “prototype” project where we focused on a single elected official. We found 18 “pathways” from the IS team to that member of Congress. Our expectation is that among those 18 pathways, there will be many individuals who can, at the least, provide IS information about that elected official, and hopefully, one or two who have some influence with that official and can help IS educate and advocate for the non-profit and philanthropic community.
Can JUTE do the same thing for my organization?
Yep. That’s what we do. Contact Sean McDonald to learn how we can help you.
*A term coined by Patricia Read, Sr. VP of Public Policy and Gov’t Affairs at IS.
You make friends and network with peers who are smart, powerful and connected people.. AND who know more smart, powerful and connected people.
(insert visualization here of networks)
Now imagine harnessing all of those networks and information.. and being able to focus on those relationships that help YOU and YOUR BUSINESS meet your goals in a simple and quick way.
That’s JUTE.
We help private, public, government and nonprofit organizations grow and become more efficient in a few simple steps.
JUTE provides valuable information about your relationships and how to leverage them. We help you grow your business and meet your goals faster than anyone else.
JUTE is a team of good people and good technology.
You meet people every single day. You make friends and network with peers who are smart, powerful and connected people.. AND who know more smart, powerful and connected people. Now imagine harnessing all of those networks and all of that information, being able to identify and focus on those relationships that help YOU and YOUR BUSINESS meet your goals in a simple and quick way.
That’s JUTE.
We help private, public, government and nonprofit organizations grow and become more efficient in a few simple steps. JUTE provides valuable information about your relationships and how to leverage them. We help you grow your business and meet your goals faster than anyone else.
Contact our team to learn more about relationship discovery and analysis services.
Learn about NEXUS, our discovery & analysis service.
I created a custom visualization for the (RED) campaign. Let me know what you think, and share it if you think it’s important to fight AIDS in Africa.
It’s World AIDS Day today. I don’t think I need to outline why this important, but if you’d like to know more, check out joinRED.com where there are videos and all kinds information. You can even pimp your Facebook page with their tools.
Something that’s really important to me (Sean) is that all kinds of traits transmit through relationships. This is what we call “sociographic” transmission, although the academicians have yet to settle on a lexicon that deems that phrase precise. We’ll leave the vocab up to Stanford…we just analyze networks.
Anyway, what I tried to reflect in this visualization is the idea that we are all influenced by our friends and that we can influence each other. This type of incluence is–obviously–one of the most powerful forces on earth.
So take a minute, do something, and convince somebody else to go (RED).
We held a free webinar for non-profits in Western North Carolina, which was well-attended. During that webinar, I explained a few things about trends in the non-profit industry, where technologies featuring social network analysis and social network visualization are being put to use by foundations and non-profits.
Using a combination of real and mocked up data, I then presented a few ways non-profits can use this type of technology–and Jute specifically–to save time and increase donation revenue.
You can watch the presentation in this video or flip through in the embedded in Google Doc below.
One of the emerging technologies that is helping non-profits identify fundraising opportunities, manage executive transitions and communicate with funding partners is social network analysis (SNA). This technology goes hand-in-hand with social network visualization (SNV). These technologies require a lot of sophisticated work with data, some training on how to read the “maps,” and how to present findings to colleagues and partners.
This webinar will be a brief overview of how some non-profits are using SNA & SNV, and a quick demo of Jute NRM Prototype, which uses these two technologies as part of a consulting process.
Learn more ahead of time
If you’d like to learn more ahead of time, watch this video from TED Talker Manuel Lima, founder of www.visualcomplexity.com — a storehouse for hundreds of types of social network visualizations.
I just read about the OpenCalais project from Thomson Reuters. The company describes it as:
Calais is a rapidly growing toolkit of capabilities that allow you to readily incorporate state-of-the-art semantic functionality within your blog, content management system, website or application.
What I interpret is that they are creating a hub for semantic data sources, so that all the formats and all the semantic options become increasingly accessible for developers. My hypothesis is that Thomson Reuters is aware that people will also want to buy data about the people, relationships and companies they access through OpenCalais and that they will be able to sell it to them.
There’s a couple intro videos–with especially well-done animations it’s worth mentioning–on their website. (Embedded below…)
There are two visualization projects in their community so far. One called Thinkpedia, using ThinkMap to visualize Wikipedia relationships and another called Wirecatch to visualize business relationships found in news stories.
Email me if you have any thoughts about how to put Calais to use.
As we move forward with Jute Beta, we’ll integrate some of the powerful network analysis and visualization tools that people have been using in Jute NRM Prototype. In an effort to promote these tools, and to help drive forward social network visualization (SNV) as a business tool, we have documented that version in video and with a features list. We will continue to maintain a version of Prototype–albeit a buggy one–for internal testing and use by people who are passionate about SNV.
I was chatting with a guy named Chris Ashford, who is a co-founder of Southern California Outdoor Adventures, a firm that organizes adventures and connects adventurers. He is also a bit of a techie and we were talking about Jute and our business network visualization interface. He made this observation (paraphrased):
It sounds like an evolution of the interface. Kinda like going from DOS to Windows.
We’ve used a lot of metaphors to explain why we believe that interactive visualization will be the interface of the future for enterprise relationship management software, but never operating systems. Spreadsheets and charts–yes. Weather data and radar imagery–yes. PC and Mac–yep. But never Dos and Windows.
But I think Chris is right. This is the metaphor. A visual interface makes it easier to absorb more information and act on it, while lowering the knowledge and experience required to adopt a piece of software. It also simplifies tasks, and can make certain functions dramatically more efficient. An operation that was very time consuming in the DOS command prompt–like moving a folder full of files to a set of new folders with the files distributed across the folders based on their contents–became easy to understand and easy to perform in Windows. What’s more, a world full of new features and functions were inspired by this new interface.
In this way, operations that are difficult to understand, information that is hard to process and act on and tasks that time-consuming and complicated in a traditional relationship management interface are made simple, intuitive and approachable in a visualization interface. Furthermore, this interface not only provides a context rich user experience, but also leads to waves of innovative thinking that results in an entirely new conception of appropriate features and functions.
So thanks, Chris, for the inspiration and the good analogy. And thanks to Matt, Dan, James and our whole team who have helped pioneer a new way of approaching interface design for business networks and enterprise relationships. I’ll write more on this soon.
Over the weekend, I sampled some code from OpenProcessing where people can share their Processing sketches, and put this together. An homage to my favorite types of visualization…
Simple controls:
r = toggle relationships (links)
b = toggle contacts (nodes)
space = change visualization mode
a = add contact (in second visualization mode)
You have to click inside the applet box to interact with it.
For those of you familiar with Processing.org, you’ll know that:
Processing is an open source programming language and environment for people who want to program images, animation, and interactions. Processing.org
I’ve been learning to use Processing in my free time, largely to prepare for a project with The Bob Moog Foundation, but it might also come in handy for work here at Jute.
Important recognition: thanks to Lorenzo Marchi and Daniel Shiffman, whose code makes up the majority of this sketch.
Recently, Google has rolled out a new set of search tools both for web search and YouTube video search. One of the featured tools is a node-link visualization engine called the Wonder Wheel. You can read more about Wonder Wheel at the Google Blog and on TechCrunch.
There’s a few things to notice about Wonder Wheel from the perspective of node / link visualization:
1) The point of centrality (the selected node) is stationary. This makes it easier to interpret.
2) It is exclusively first and second degree relationships that are shown. There’s no option to see search terms related to one-another by three or four degrees, let alone the famous six degrees.
3) There is only one piece of information presented per node.
These are just a few interesting things. Google’s persistent focus on adoption has clearly driven the design decisions. Given their traffic and their clout, more people have probably tried Wonder Wheel than have used all of the other node / link visualization applications put together. Overall, it’s very simple, not too powerful, but highly approachable. I hope they continue to advance this project and roll visualization-as-the-interface into more of their projects. Great work, Google.
