InfoVis 2003 Contest
History of Updates
In general: only minor corrections to the data have been made to correct problems you/we find. No-one suggested any changes or clarification to the task descriptions. Overall we had very few comments and questions. The data format never changed so that people could start working early.
Thanks to Michael Spenke and Christian Beilken for finding the problems and to Cyndy Parr and Bongshin Lee for fixing them. This process makes it obvious how hard it is to create those classifications without a good mean to look at them as some of the errors had been in the dataset for years.
The zip file with all the datasets has been updated as well, and is now called iv03contest_data_03-04-16.zip
Corrections in both trees:
Inconsistent data in line 45250:Correction in tree B:
The path from the Suborder 'Eucladocera' to the root contained another node of rank Suborder, namely 'Cladocera';
Inconsistent data in line 45471:
The path from the Order 'Conchostraca' to the root contained another node of rank Order, namely 'Diplostraca'
Here the problem was in our source data. To solve both problems, we changed the ranks to be Superorder for Diplostraca and Order for Cladocera. This is consistent with at least one classification by an invertebrate anatomist: http://www.lander.edu/rsfox/invertanat.html
Inconsistent data in line 589395:All the other problems resulted from misconfiguration of the crocodile branch. We put Crocodiles where they belonged and eliminated some nodes.
The path from the Genus 'Geophaps' to the root contained another node of rank Genus, namely 'Columba';
Geophaps remains a genus but has parent Columbidae instead of Columba.