Why ACL reviewing sucks (not as much anymore ...)

 

Sounds ok, but ...

Allright, no support for web technologies, then. But luckily, there's an alternative ...

Really, file by file ?

For me, this means running the upload dialog some 88 times (excluding monographies and pre-2013 stuff):

 Rather not ...

Let's try aggregator portals ...

Nice, DBLP works, but only for content not mediated by Mendeley ... This includes all ACL publications:

A few papers at workshops with CEUR proceedings go through, but this is not representative
Using the ACL anthology itself works much better, but this also covering only 1/3 of my publications -- obviously the more relevant ones in this context, but still ...

Let's see if I get a response to the following email:

 

Update 2020: I wrote the above as a result of a really frustrating evening, and it pretty much summarized my feelings at the time, because the burden of automated paper assignment was way too much put on the reviewer. Effectively, if you did this properly, this was probably more effort than just doing bidding over the full set of submissions (which was around a few thousand at the time). (BTW: As for the requested response, I never got one.) However, the situation improved, and while just confirming to review for ACL-IJCNLP, I figured, I should reflect this. Nowadays, information on previous papers is automatically drawn from aggregator portals (as it should be). With this, the administrative workload for setting up a reviewer profile has returned to normal standards. However, I chose not to change the title of the post completely, but rather add the "not as much" part, because it would be a completely different post then. Of course, there are things that can be criticized about peer review in general, but ACL is no worse than anyone else.