Welcome to the Wiki for T-REX 2010
Announcing T-REX 2010
TADA (the Text Analysis Developers' Alliance) and Duquesne University welcome your participation in T-REX 2010, the TADA Research Evaluation eXchange.
Following the success of
T-REX 2008 at Digital Humanities 2009 (University of Maryland), we are happy to continue the tradition of a multifaceted, multicategory "best ideas" forum for the development and exchange of ideas in the field of "text analysis" (broadly defined). We invite all interested participants to share in the next round and to help us expand our offerings.
The only hard and fast rule at this point is : ''All participants must participate!''
What is T-REX?
T-REX is modelled after both competitions (such as TREC) as well as less formal idea exchanges(such as
MIREX). We hope to stimulate creative competition and the exchange of ideas about questions such as what we can do with texts, how to do it best, and why we would want to do it in the first place.
To do this, we hope to generate a set of collaborative "challenges" to which participants will submit ideas/solutions. All ideas will be "judged" by a panel of judges (to be determined); this judging may involve the actual running of code provided by participants or peer-review by volunteer experts. Entries judged to be the best will receive valuable Pittsburgh-themed prizes (such as bottles of Heinz products).
More importantly, we encourage all participants to join us at Digital Humanities 2010 (London) or a similar venue TBD, for the "Research Evaluation eXchange" part of this process, where we will share our ideas, evaluate them, discuss them jointly, and if possible improve them. In this process, while there may be winners, ''there are no losers.''
How to participate?
- Send email to ''trex dot challenge atsign gmail dot com'' with your name and contact information.
- Edit this wiki with your ideas for interesting challenge problems. These can range from abstract questions like "what's the best way to read text?" to hard-core technical issues such as "what's the fastest way to find all prepositions in an English document?"
- Help refine proposed problems, including suggesting judges or judging methods, appropriate datasets, and so forth.
- Participate by submitting solutions to the challenges by the appropriate deadline
- Come to DH2010 to collect your well-earned kudos.
Problems
As with MIREX, we will judge any proposed challenge problem for which we get enough entrants. Also as with MIREX, the list of problems below is not fixed; feel free to propose something, and be as general or as specific as you like.
- Best New Tool
- Best Idea For a New Tool ?
- Best Idea for Improving an Existing Tool ?
- Best Experiment in Text Analysis ?
Detailed rules
... will be forthcoming.
Timetable
... will be forthcoming.
Sponsors
... will be forthcoming.
--
PatrickJuola - 29 Jul 2009
