Response to the invitation for comments
Posted by: eoswald
Date: 08 August 2013 20:50
I welcome an open discussion about the publication routes that we currently have in our community and the option to comment on the current proposal. I find myself in agreement with several of the points raised, but at the same time there are two key points which I dislike and want to be different. To explain what these are I require several lines of text so please bear with me.
To begin with, I think it is important to be clear on goals, very much alike how the proposal approaches the problem. Personally I think that our goals could be stated more simply:
a) we want enjoyable conferences which are inclusive (i.e. all community should have a reason to go there), albeit there is certainly a need for specialist workshops/conferences
b) reviewing needs to be credible and ethical and we need to be clear what our criteria are (i.e. are we selecting based on correctness, completeness, or potential to attract lots of citations only, etc.)
c) we want to reduce unnecessary reviews and overhead so that a) and b) can actually materialise.
In light of these goals, and the clear inefficiency of the hidden multi-round reviews that we are doing now, I completely agree that we need to finally drop the idea of relying on conferences as major publication venues.
However, I don't find the idea of a single journal that essentially hoovers up all conference and workshop publications enticing, and the idea that conference chairs select only/mainly(?) from already published work doesn't sound good to me either.
I would very much advocate that we think about setting up 'Transactions on' (using IEEE terminology) or 'Communications' (ACM terminology) in the already identified specialist areas. This is not only to achieve some differentiation of topics, but also different communities might want to have slightly different review processes, and this could be addressed in this way. For instance, among practitioners the idea of a first 'quick' round and subsequent (almost) open-ended multi-round review process finds some supporters. Joan Damen's post in this forum echoes this as well. This might be because implementation or experimental work can take many different routes and so a decision about the work can be based on it's applicability per se but the precise nature of the implementation/experimentation is up for discussion and some paper benefit greatly from reviewer comments. This is at least an experience from the many CHES PCs I have been on over the last decade.
By having some more journals available conferences would be freed from having to ask for original research only. We could hence have conferences with some original contributions without proceedings but which invite an actual conversation about on-going work! Conferences could hence become much more than just the 'one talk after another' event, they could have interactive sessions, overview talks, etc.
Obviously, the whole community takes a gamble by initiating any kind of radical change in the publication model as any new journal will start from zero in terms of impact. Because impact is important in the long run we better make this change sooner rather than later, as LNCS is a poor currency. And this sad fact is particularly concerning for young researchers in academia whose potentially great results will be undervalued in the wider academic community just because they appear as LNCS proceedings.
Thanks to all those who drafted the current proposal and thanks to everyone who read my lengthy ramblings until this end.