| 
  • If you are citizen of an European Union member nation, you may not use this service unless you are at least 16 years old.

  • You already know Dokkio is an AI-powered assistant to organize & manage your digital files & messages. Very soon, Dokkio will support Outlook as well as One Drive. Check it out today!

View
 

Research Report by Claire Ihlendorf

This version was saved 11 years, 10 months ago View current version     Page history
Saved by ihlendorf@...
on May 17, 2012 at 10:52:06 am
 

 

Research Report: History Flow tool 

 

By Claire Ihlendorf, Playful Visualizations at Work, Working visualizations at play Team

 

1. Abstract. History Flow tool for visualizing the evolution of collaborative documents.  It creates a colorful and intricate visualization that best represents Wikipedia or MoinMoin pages.  It is a downloadable tool that free and relatively easy to use.  The utility of a tool helps users see the collaborative patterns of editors to these pages.  The tool can also create visualizations for directories of text.  It was created by Fernanda Viégas and Martin Wattenburg in 2003 for IBM. 

 

2. Description.  History flow is a visualization tool that tracks the evolution on Wiki sites or on a directory of texts.  It was created by Fernanda B. Viégas and Martin Wattenberg for IBM in 2003.  Both Viégas and Wattenberg are recognized leaders in data visualization and have collaborated on many projects including the 2007 visualization website, Many Eyes (also for IBM).  History flow was initially created to examine the website Wikipedia which was a new phenomenon at the time.  The visualizations that the history flow tool would create helped Viégas and Wattenberg notice collaborative patterns and help suggest the implications of these patterns.  The results of this study were published in a paper titled; “Studying Cooperation and Conflict between Authors with history flow Visualizations” presented at CHI 2004, in Vienna.    

 

The program is downloadable for free from IBM and was written in Java 1.4.  Once it is downloaded, a user is able to visualize the history flow of a directory of text documents.  There are example directories preloaded to the program to test different directories (either with an index or without).  In order visualize the evolution of a Wikipedia page, it is necessary to download an older version of Jtidy (an HTML syntax checker).  This additional download is available from wiki.digitalmethods.net and is used to export the catalogued changes and revisions from a specific Wikipedia page.  It allows a user to be able to import Mediawiki or MoinMoin pages.  It also makes the user interface much easier for the Wikipedia visualization; one needs only input the page title in the prompt:

 

Additionally, it is easy to change the language in order to search the Wikipedia pages that are not written in English.  To change the search from English to Spanish as shown in the image above,, the user needs only change the .en to .es (the same would work for other corresponding country codes). 

 

The visualization that is created appears weaved together.  Each contributor has a specific color and each version has its own line.  Thus, when a contributor adds or deletes, a section, his or her color is continued on the line.  The weave is made when the colors are continued horizontally.  Black gaps represent sections that have been deleted.

 

The history flow visualizations can be viewed in a few different options.  The visualizations can be organized by version or date and also by the collaborative authors, an individual author, recent changes or by and age view (a gray-scale visualization that notes the “age” of contributions; the darker the color, the older the contribution).

 

3. Commentary.  For its purposes, the tool is very effective.  It examines and documents the changes made to Wikipedia pages very well and the resulting visualization is both colorful and useful.  Since a person who visits a Wikipedia page might not take all of the edits into consideration, this visualization tool helps demonstrate the amount of collaborative work that goes into this type of page.  Additionally, it is a good representation of the history (or as the authors call it, “the evolution”) of a page.  An interesting result is to examine controversial topics and see the vandalism that occurs.  In these cases, there is a large black gap where the content of the page has been erased.  This type of vandalism occurs most often on controversial pages, but also can exist on random pages.  When I created a visualization of the Wikipedia page on Wikipedia, there were five clear instances of vandalism:

 

One limitation that I have been experiencing with recent visualizations is an error message if the page title is not input correctly.  I did a number of searches and certain pages did not show up; however, when I capitalized the names correctly, they visualizations appeared.  As such, I believe the Jtidy download is case sensitive. 

 

For the purposes of our project this quarter, I was looking for a tool that created a pretty visualization, and not necessarily a useful one.  I was attracted to the history flows because of their use of color and complex pattern.  I found these visualizations interpretively useful when I considered the study of the evolution of Wikipedia pages.  However, when examining a text directory, I did not find the tool as helpful.  This aspect of the history flows did not bother me as I am more interested in the image it creates and not if said image will help me understand the work better. 

 

I ran into a few problems in attempting to create visualizations based on the text with which I am working.  The main issue is the lack of collaborative authors on the texts.  However, when I input text directories filled with files of my own work (documents relating to my dissertation) the result was better than anticipated, and a colorful history flow image was created.  I plan to continue working with the tool to test its limit and see if I can manipulate either the tool or my data so that an optimal result is achieved. 

 

4. Resources for Further Study

 

IBM history flow site

 

Fernanda Viégas’ personal website

 

Martin Wattenberg’s personal website

 

"Between Authors with History Flow Visualizations."  Paper presented at CHI 2004, Vienna, August 2004.  Proceedings of CHI Conference, April 24-29, 2004, Vienna, Austria), CHI 6(1): 575-82.

 

To download history flow

 

A useful ‘how-to’

 

 

 

Comments (0)

You don't have permission to comment on this page.