- From: Charles McCathieNevile <charles@sidar.org>
- Date: Fri, 18 Feb 2005 13:41:07 +0100
- To: "Myriam Arrue" <myriam@si.ehu.es>, public-wai-ert@w3.org
Aupa Myriam! Hi folks In fact Fundacion Sidar is working on just such a use case. We have a tool (Hera) which uses the results of a number of different evaluations - some automatic done by the tool, some manual done by a person, some automatic and done by an external tool. So a basic use case is to combine these results. In order to help us in developing the tools, we also compare results to see if a new tester or automated test gives the same results as a tester we trust. Cheers Chaals On Fri, 18 Feb 2005 12:33:28 +0100, Myriam Arrue <myriam@si.ehu.es> wrote: > Hi everybody! > I'm Myriam Arrue from the Laboratory of Human-Computer Interaction in > the University of the Basque Country. I'd like to start the discussion > about the scenarios of EARL. > One of the main objective of EARL is to combine different tools' > evaluation results in order to compare them. Another important feature > of EARL is that it can be used for exchanging data between tools. > In my opinion, these two objectives can be integrated in a scenario in > order to clearly describe to evaluation tools developers the advantages > of using EARL as the errors reporting format. > A scenario where a tool or software application invokes different > evaluation tools combining the different results obtained in EARL in one > complete evaluation result report can be useful for this purpose. This > scenario would highlight the need of interaction between evaluation > tools and also the importance (as described in Evaluation Suite > (http://www.w3.org/WAI/eval/ ) of evaluating the web content with at > least two different evaluation tools. > Waiting for your opinions, > Myriam > > -- Charles McCathieNevile - Vice Presidente - Fundacion Sidar charles@sidar.org http://www.sidar.org (chaals is available for consulting at the moment)
Received on Friday, 18 February 2005 12:49:39 UTC