Publishing conference proceedings and other kinds of edited collections as special issues of journals has a long history. But lately (partly as a reaction to perceived shortcomings of the more traditional CS system of publishing a preliminary version of a paper in a conference and then a full version in a journal) there's been increasing pressure to do this for more conferences. Which raises the question: how are we supposed to format these things in our bibliographies and bibtex files?

Some standard but not-very-satisfactory solutions include: pretending the journal is a book series and using the series= parameter of an @inproceedings bibtex entry (but where does the journal issue number go?); formatting it as a normal conference proceedings paper and omitting the journal name and volume/issue data; formatting it as a normal journal paper and omitting the conference name and editors; or putting the data that doesn't fit into a note, formatted manually, and forgoing the uniform and flexible formatting provided by bibtex.

I recently ran across an intriguing alternative used for some entries in MathSciNet: use @inproceedings, but with the journal= parameter giving the journal name (in place of series=), and with volume= and number= parameters again used as you would for a journal paper. Here's an example (not the one from MathSciNet):

@inproceedings{FAvKSCM,
title = {Meta-representation of shape families},
author = {Fish, Noa and Averkiou, Melinos and van Kaick, Oliver and Sorkine-Hornung, Olga and Cohen-Or, Daniel and Mitra, Niloy J.},
booktitle = {Proceedings of ACM SIGGRAPH 2014},
doi = {10.1145/2601097.2601185},
issn = {0730-0301},
journal = {ACM Transactions on Graphics},
month = Jul,
number = {4},
pages = {34:1--34:11},
volume = {33},
year = {2014}}

If that worked, I'd expect it to be formatted something like:

• Fish, Noa; Averkiou, Melinos; van Kaick, Oliver; Sorkine-Hornung, Olga; Cohen-Or, Daniel; Mitra, Niloy J. (July 2014), "Meta-representation of shape families", in Finkelstein, Adam, ed., Proceedings of ACM SIGGRAPH 2014, ACM Transactions on Graphics 33 (4): 34:1–34:11, doi:10.1145/2601097.2601185

But sadly, it doesn't work with any of the standard bibliography styles I tried it on (abbrv, splncs, and amsplain). In all cases it gives a warning and then produces a citation that is missing the journal name and issue number.

Pretending the journal is a book series and using series= instead of journal= worked significantly better, although not perfectly. From abbrv, splncs, and amsplain I got respectively

• N. Fish, M. Averkiou, O. van Kaick, O. Sorkine-Hornung, D. Cohen-Or, and N. J. Mitra. Meta-representation of shape families. In A. Finkelstein, editor, Proceedings of ACM SIGGRAPH 2014, volume 33 of ACM Transactions on Graphics, pages 34:1–34:11, July 2014.

• Fish, N., Averkiou, M., van Kaick, O., Sorkine-Hornung, O., Cohen-Or, D., Mitra, N.J.: Meta-representation of shape families. In Finkelstein, A., ed.: Proceedings of ACM SIGGRAPH 2014. Volume 33 of ACM Transactions on Graphics. (July 2014) 34:1–34:11

and

• Noa Fish, Melinos Averkiou, Oliver van Kaick, Olga Sorkine-Hornung, Daniel Cohen-Or, and Niloy J. Mitra, Meta-representation of shape families, Proceedings of ACM SIGGRAPH 2014 (Adam Finkelstein, ed.), vol. 33, ACM Transactions on Graphics, no. 4, July 2014, pp. 34:1–34:11.

(with a warning from both abbrv and splncs about using both volume and number together).

I don't normally use biblatex, but I also tried it (with its default settings), with similar results. It didn't complain about journal=, but it didn't show the journal name in the reference when I used it. With series=, I got the output

• Noa Fish et al. “Meta-representation of shape families”. In: Proceedings of ACM SIGGRAPH 2014. Ed. by Adam Finkelstein. Vol. 33. ACM Transactions on Graphics 4. July 2014, 34:1–34:11. DOI: 10.1145/2601097.2601185.

which is a bit mixed up, but at least has all the information visible. If you do plan on using biblatex only, there is a better option available through its additional fields:

@article{FAvKSCM,
title = {Meta-representation of shape families},
author = {Fish, Noa and Averkiou, Melinos and van Kaick, Oliver and Sorkine-Hornung, Olga and Cohen-Or, Daniel and Mitra, Niloy J.},
issuetitle = {Proceedings of ACM SIGGRAPH 2014},
doi = {10.1145/2601097.2601185},
year = {2014}}
So it looks like, for now, despite its shortcomings, series= is the way to go. But biblatex is promising and might (with styles that better match publisher requirements) be a better eventual solution.