PAN 2015 Author Profiling Software Argument Format Question

[arspou...@sheffield.ac.uk]

Hi, 

I'm a little confused as to what arguments my software should be taking when submitted to TIRA.

In the submission guidelines, it tells me to create a training and a testing software:

> myTrainingSoftware -i path/to/training/corpus -o path/to/output/directory

> myTestingSoftware -i path/to/test/corpus -m path/to/classification/model -o path/to/output/directory

Yet the TIRA software runs are of the form:

> mySoftware -c $inputDataset -r $inputRun -o $outputDir

Am i correct in assuming that  -c $inputDataset maps to -i path/to/training/corpus, -r $inputRun to -i path/to/test/corpus?

Thanks,

Adam

[Martin Potthast]

Hi Adam,

sorry for the inconsistency between PAN's web page and TIRA. I'll fix
that as soon as possible.

> I'm a little confused as to what arguments my software should be taking when
> submitted to TIRA.
>
> In the submission guidelines, it tells me to create a training and a testing
> software:
>
>> myTrainingSoftware -i path/to/training/corpus -o path/to/output/directory
>
>> myTestingSoftware -i path/to/test/corpus -m path/to/classification/model
>> -o path/to/output/directory
>
> Yet the TIRA software runs are of the form:
>
>> mySoftware -c $inputDataset -r $inputRun -o $outputDir

All of these are just examples. The only important thing about them is
the input parameters themselves, not the names of the switches.

The training software should take two parameters, which are paths to
the input dataset and the output directory. In TIRA, these paths are
encoded as variables $inputDataset and $outputDir, since they are
dynamically filled in when you hit "Run".

The testing software should take three parameters, which are paths to
the input datasets, the output directory, and the model which was
trained using the training software. The model is the result of
running the training software, i.e., a run of the training software.
In TIRA, these paths are encoded again as variables $inputDataset and
$outputDir, and in the case of the model the variable is called
$inputRun, i.e., the previous run of the training software.

Best,
Martin

PS: We're working on something more easy to follow, but it'll take
some more time...

--
Dr. Martin Potthast
Bauhaus-Universität Weimar
Digital Bauhaus Lab
Bauhausstr. 9a
99423 Weimar
Germany

+49 3643 58 3567
+49 171 809 1945

www.potthast.net