System description papers

[Felix Stahlberg]

Hi,

do we have to anonymize system description papers? The website seems to indicate that but afaik it is not common practice for system descriptions (see eg. WMT), and I find it a little difficult to do that - it might result in multiple "The [blank]'s Grammatical Error Correction Systems for BEA-2019" papers if people are as uncreative as I am with paper titles, and it would not be possible to refer to other submissions.

Best, Felix

[BEA 2019 Shared Task]

Hi Felix,

Another participant actually asked the same question in a private email so it's probably worth telling the whole group.

Yes, the paper authors should be anonymised, but no, you shouldn't anonymise team names or other identifying information. While this will make it relatively easy to identify specific authors, all shared task papers will get published anyway (unless they are low quality) and so reviewer biases don't really matter. Shared task paper "reviews" are ultimately more about proofreading and feedback, and so lack of anonymity won't be held against you.

We only ask for author anonymisation because the main BEA organisers told us it would make it easier for them to handle the review process if everything is submitted in the same way.

Hope that helps,

Chris

--
You received this message because you are subscribed to the Google Groups "BEA 2019 Shared Task: Grammatical Error Correction" group.
To unsubscribe from this group and stop receiving emails from it, send an email to bea2019st+...@googlegroups.com.
To post to this group, send email to bea2...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/bea2019st/aac8bf9c-2936-48e7-a059-3f2a2df13752%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.