Seeking contractor to work on NodeXL project

May 31, 2011 at 10:59 PM

Our team is kicking off an innovation challenge (a “digital jam”) which will use a collaboration software called BrightIdea to “crowdsource” and select innovative product ideas around some new areas of our business.  The primary target participants will be our R&D engineers.  We are planning to run a series of such events over the next 12-18mos covering various areas.

 As we have moved closer to the launch date of our first jam, we have become interested in visualizing the activity and interconnections that occur within the event platform.

Please let me know if you or someone you know is interested in a consulting/contractor engagement to assist with the visualization aspect of this project.

Thanks!

-ng

Jun 1, 2011 at 2:36 AM

Hi Nicholas,

Do you have any ideas about the shape of the data set that you have to play with? Usually the biggest obstacle in these organizational knowledge projects is finding ways to collect data that is unobtrusive, fairly comprehensive and valid.  Visualising it is the easy part.

Thanks.

Jun 1, 2011 at 1:13 PM

We will have access to the data, but not exactly sure of the format yet.  It is likely that we will have to perform some transformation(s) before it is ready for NodeXL (or another app), and there may be some other data that we would fuse into this before visualizing.   How significant this transformation effort will be is a key unknown at this point.....and understanding the optimal shape of the data for purposes of visualization (via NodeXL or otherwise) is a big part of it.

Coordinator
Jun 1, 2011 at 2:09 PM

NodeXL imports a variety of data file formats: GraphML, CSV, UCINet DL files, Pajek .Net files and direct import from sources like Twitter, Flickr, YouTube, the Web, email, and more.

If you are creating your own data set, the simplest format is the "edge list" which is a set of pairs of names or identfiers, optionally with attributes about each identified entity and attributes of the relationship.

For example, if a group of people sent each other emails we might create an edge list that looks like:

A, B, 6

B, C, 2

C, A, 1

This describes a "triangle" with different "weights" for each edge or tie.

More sophisticated approaches can add more attributes to describe the entities and their relationship.  For example, GraphML allows for more detailed information about both edges and vertices to be directly imported into NodeXL.

A detailed description of the GraphML format is available here: http://graphml.graphdrawing.org/

Regards,

Marc