Jump to content

Talk:List of computer science conferences

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Rank-based listing

[edit]

The categories, if not the conferences themselves, need to be alphabetized. Personally I disagree with basing the order of the conferences on their "ranking", as this list serves to index the Wikipedia, not to publish a Top 10. In addition, if the rankings were established using subjective and non-obvious criteria, the order of the list may be a copyright violation. And finally, the list may contain entries not included in the source. Ham Pastrami (talk) 23:56, 15 March 2008 (UTC)[reply]

I'm doing research aimed at speeding up computer simulations for large power systems. I'm very happy with the ranking system, it has saved me me a lot of time, and it will probably continue to do so. —Preceding unsigned comment added by 132.207.137.220 (talk) 15:18, 29 May 2008 (UTC)[reply]

I disagree with alphabetizing. The sources (Citeseer and MSRA's Libra) used as a general ranking guidelines are relatively reliable. I repeat, they are only used as a general guideline and one with objective opinion should feel free to change the order.

Wikipedia should be used as a source of information, and the importance/impact of the listed conferences is such information. We have the opportunity here to actually provide a non-biased impact-factor-based listing of those conferences and leverage the opinions of the community to reach a general consensus. What help would this list be to a Computer Scientist if it lists unknown/low-impact conferences in the same order as highly prestigious? There is absolutely no copyright violation, this claim is ridiculous.

Of course the listing contains entries not included in the source, we are not copying citeseer's and libra's listing which may be outdated. If someone knows about a non-indexed conference, she should feel comfortable to add it, as long as it is sufficiently reputable and most importantly academic.

But neither of those lists arrange conferences by category, which this article does do. Since you freely acknowledge that additional items, which do not appear in the sources, are added to this list, how then do you propose to order them consistently and objectively? First of all, the list should be pruned so as to contain only noteworthy items, as is policy at WP:SAL. Indeed, SAL also provides examples of list formats, of which alphabetization is the first, and "rank of importance" is not even mentioned. Pruning the list solves any question of how "useful" the list is. Wikipedia does not judge importance, only notability. Either a conference is notable enough to carry an article or it isn't. If you want to provide a list that allows lookup by importance, it already exists at the source, which can be linked to externally. There's no need to copy and paste it here. Do you have an actual policy argument in favor of rank-based ordering? Ham Pastrami (talk) 10:16, 16 March 2008 (UTC)[reply]
This list contains only noteworthy conferences. It is up to us to clean it from conference spamming that occasionally takes place. AFAIK all listed conferences are included in citeseer and libra, however it is possible that this will change in the near future as new noteworthy conferences are created. It is up to the community (knowledgible CS researchers) to place them in the list according to their perceived quality. Citeseer's and libra's ranking are only used as a general guideline, to ensure that low-cited conferences are not listed up there with the most prestigious one. This is why the ranking is described as rough and non-authoritative
For the record, here is the determination of rank from Citeseer: Impact is estimated using the average citation rate, where citations are normalized using the average citation rate for all articles in a given year, and transformed using ln (n+1) where n is the number of citations. I don't know if that measurement qualifies as an obvious fact or not. I'm not a copyright lawyer. Are you? In any case, it's just a possible issue with any copy-paste list, but this is on top of all the other problems that this list already has. Ham Pastrami (talk) 10:35, 16 March 2008 (UTC)[reply]
Citeseer's method is not obvious neither is Libra's (which is much more complicated btw). This does not mean we cannot use them as a guideline. The fact that they do not categorize per field and that we do not exactly follow their order protects wikipedia from copyright violation. Besides this article heavily cites them and offers more visibility to them. If they have a problem with that let them complain first, but I am certain this won't be the case. Using them as a guideline means that you won't see a conference that has 0.001 impact factor or 1/100 citation ratio listed among the top in a field. So from Libra all we use is the citation ratio of each conference, which is publicly available, non-secret and non-copyrighted information. If this was copyrighted, Libra and Citeseer would be infringing on the copyrights of ACM, IEEE etc just because they analyze their conference' statistics. I really do not understand why you insist on removing useful information from an article based on invalid copyright concerns.
Is this because you would rather the readers have no valid references and have to go to the likes of conferencebay.com instead? This is a computer science article, not an opportunity for spammers to advertise generic search engines. If you are just a conferencebay.com's affiliate, please just leave this article as is to be handled by real computer scientists.

More about listing

[edit]

Quote from the article "ranking of their quality and impact based on ...".


  • libra and citeseer (did't check Harzing) use different "metrics" to rank conferences. Which metric is used in the article? Your own? Could you give it to us?
Libra and citeseer are used as a general measure of the citation-based impact of a conference. They are used only to distinguish high impact conferences from low impact conferences. High impact conferences are almost always high quality conferences. Nevertheless citation-based ranking has several flaws and cannot be accurate. For example one cannot conclusively infer that the #1 cited conference is an overall better conference than the #2 or #3. One can infer however that a conference is better than another if it has 3-10 times higher citation impact factor.
Given this rough ranking the community is able to rearrange the ordering of conferences as it feels appropriate. This will lead to an informed community-driven conference ranking. Still, this ranking should not be taken at face value, i.e. using the example you mention below, the ordering should not be used to infer that POPL is a better conference than PLDI or vice versa. However, one could use it to infer that POPL is a better conference than FLOPS, which is in agreement with the vast majority of PL researchers' view.
  • how are the conferences chosen for this article? libra's top three ranking on "software engineering and programming languages" is

1. POPL 2. PLDI 3. ECOOP

For instance, I don't see ECOOP in the article.

You should feel free to add ECOOP and improve this article, it was an inadvertent omission
  • Is listing conferences in their fields a good idea? Many conferences are multi disciplinary. Look ASPLOS - International Conference on Architectural Support for Programming Languages and Operating Systems. You could list it under "programming languages", "operating systems" or "computer architecture". By listing such conferences in one field distorts the picture of the other fields.


71.131.196.254 (talk) 20:35, 20 May 2008 (UTC)[reply]

Update: I've just seen that ASPLOS like 32 other conferences, are listed in several fields, but there are other conferences like CASES - Compilers, Architecture, and Synthesis for Embedded Systems which could/should be listed in several fields as well.

Please feel free to improve this article by listing other multi-disciplinary conferences in their respective additional fields
We have decided to list conferences in only one field. We compensate by adding in each list the "See also *". It is well known that many fields partially overlap, e.g. networkin, distributed computing and OS, so people who are interested in finding good confs in this general area can look up good confs in all the overlapping fields. John.legal (talk) 16:02, 27 January 2010 (UTC)[reply]

71.131.196.254 (talk) —Preceding comment was added at 22:38, 20 May 2008 (UTC)[reply]

Table View

[edit]

Wouldn't presenting the list of conferences in a table view that also lists the important facts like: submission deadline, notification deadline, and link to accepted papers --- be much more useful than just a huge list? —Preceding unsigned comment added by Tirvus (talkcontribs) 12:31, 17 July 2008 (UTC)[reply]

It would be useful. But only if it maintains the current field taxonomy and the quality-based approximate ranking. The previous attempt to add a table was nothing else by a blatant copy paste of previous conference statistics pages.

The bottom line is do not delete existing information, just add new one and it would be fine. Instead of a link to accepted papers, perhaps it is better to add a link to the latest web site of the conference. Submission, notification etc deadlines is a very useful addition, but it should be updated yearly. —Preceding unsigned comment added by 128.195.220.163 (talk) 20:12, 26 August 2008 (UTC)[reply]

Instead of trying to maintain submission deadlines, etc., here, you can add links to confsearch.org, some examples: [1] [2]. For example, there could be one such link per section. — Miym (talk) 20:43, 10 January 2009 (UTC)[reply]
this is a great idea. confsearch.com is an excellent source of conference info. —Preceding unsigned comment added by 66.57.47.124 (talk) 21:04, 18 January 2009 (UTC)[reply]
[edit]

What is the preferred thing to do with the conferences that are incorrect blue links? For example, SPAA is a blue link to something completely unrelated. As another example, GD is a link to a disambiguation page (that does not even mention the conference). Until we have an article on the SPAA conference in Wikipedia, should we (1) change "SPAA" to a red link "SPAA", or (2) change "SPAA" to "SPAA", or (3) do nothing? The current list with a lot of blue links seems to give the wrong idea of the coverage of these conferences in Wikipedia. — Miym (talk) 17:43, 18 January 2009 (UTC)[reply]

(1) and (2) appear to be preferable, but there is no need to be too strict about that. I noticed that people are slowly fixing those links.

List of distributed computing conferences

[edit]

I have been pondering how to improve this list; in particular, how to make sure that the conferences listed here are notable, and that (most) red links are really worth an article. As an experiment, I tried to tackle the section on distributed and parallel computing. In the end, it seemed best to create a separate list for the section; here is the result: list of distributed computing conferences (it turned out to be surprisingly hard to find reliable sources for something like this). Feel free to edit, merge, kill, or whatever you think is best. — Miym (talk) 00:53, 21 January 2009 (UTC)[reply]

This is a very good effort. I guess your approach is to include in your main list only the most notable conferences in each field

and to justify their inclusion based on references in their quality. I would caution about mentioning the australian core ranking. It is completely off for many conferences and journals, e.g., listing Usenix Security as B conference. The australian ranking as well as the other random rankings one can find by googling "conference ranking" are everything a ranking should not be. Biased to a high degree, based on the wrong criteria and not flexible. At least this article bases the ranking on quantifiable metrics (impact factors etc) and can be edited in a responsive way.

I believe the approach of the current article and your list per individual category can be complementary. The current list would include all known non-fake conferences and attempt to rank them, while your list would include only the most notable ones per field. —Preceding unsigned comment added by 66.57.47.124 (talk) 22:06, 21 January 2009 (UTC)[reply]

Thanks for your comments! Yes, my idea was to include only notable conferences. I am aware that the rankings are biased. That's the reason why I am only using them as an additional positive evidence of a notability of a conference – just like I have used some fairly random and biased lists in textbooks as a positive evidence of notability. In the list that I composed, (1) no conference is excluded because it got a low ranking – it is ok to add low-ranked conferences as long as there is some other evidence of notability – and (2) no conference is included only based on a ranking – some other independent evidence of notability is required. — Miym (talk) 23:17, 21 January 2009 (UTC)[reply]

Clean-up?

[edit]

WP:SAL says that "Each entry on a list should have its own non-redirect article in English Wikipedia, but this is not required if the entry is verifiably a member of the listed group, and it is reasonable to expect an article could be forthcoming in the future." So can we simply remove all non-notable red links from this list? And require that when people add new red links, they also provide references that establish the notability of the conference? — Miym (talk) 15:56, 9 February 2009 (UTC)[reply]

Core Ranking

[edit]

The core ranking is arbitrary. Wouldn't it be better to completely ignore it and only rely on citeseerx libra and other more reasonable sources? The same applies for pretty much all the other conference rankings out there, especially the conferenceranking.org/net sites. —Preceding unsigned comment added by Nick.dorvas (talkcontribs) 21:41, 2 June 2009 (UTC)[reply]

This list is using Core ranking only as a positive indicator of notability. Having an "A+" or "A" ranking in the Core list seems to be a reasonable indicator of a notable conference. I have not seen a single fake conference with such rankings, have you? (This does not mean that a non-existing or low Core ranking implies that the conference is non-notable. You just need to find some other sources.) — Miym (talk) 22:21, 2 June 2009 (UTC)[reply]
I would tend to agree with you if it was not for the core ranking listing as A and A+ conferences some really non-influential 3rd tier ones. Take for example ICIS, it has 1000 papers and 1500 citations... citeseer does not even list it. I don't know anyone who has ever considered publishing there. As I said earlier, at least citeseerx and libra do not claim authoritative results (they are automated after all) and there it is extremely unlikely that a bad conference can stack up many citations and rank very high. So back to the point, I agree with you on that we need to use these rankings as indicators of notability and not as indicators of non-notability. Citeseerx and libra achieve this. On the other hand the core ranking fails both on reliably identifying notable venues and reliably identifying low tier ones.
Interesting, I haven't spotted any extremely bad conferences on A+ and A tiers. Are you sure that ICIS is dubious? I do not know the field of information systems at all; which are the top conferences of the field? There seem to be other rankings [3] [4] that put ICIS among other top-ranking conferences, and this blog entry gives the impression that ICIS is a highly selective venue. And of course there is the Wikipedia article, too, which claims that it is the "most prestigious" gathering. You might want to mention your concerns on Talk:International Conference on Information Systems, if you think it is not notable enough? — Miym (talk) 19:52, 3 June 2009 (UTC)[reply]
ICIS is definitely not a top CS conference. There should not be a separate IS ranking, neither to have this conference rank right up there next to SIGMOD or VLDB. —Preceding unsigned comment added by Nick.dorvas (talkcontribs) 23:01, 3 June 2009 (UTC)[reply]
On another more philosophical note, it is is true that the Core ranking does get some (or most) venues right, however their mistakes are glaring and are likely intentional. For example they list NSDI and OSDI as a tier A instead of A+ conference, and USENIX Security as a B conference! At the same time they have much less influential conferences such as ICDCS listed as A+. For other examples of their bias or incompetency see https://mailhost.cs.mu.oz.au/pipermail/core-csprofs/2007-February/000089.html . It would be better if this article does not give them even more visibility than they are already enjoying due to their high pagerank. It is really amazing how much more reliable this list is compared to the Core ranking. Especially in light of the fact that the Core ranking was done using Australian tax payer money, while this wiki list is done with the work of a few volunteers. —Preceding unsigned comment added by john.legal contribs) 16:53, 3 June 2009 (UTC)[reply]
I think the ideal solution would be to find more reliable references, so there would be less need to use any of these rankings. At least for truly notable conferences, better references exists – see, e.g., the 2nd paragraph of STOC or the textbook references in PODC – they are just sometimes very difficult to find. If anyone happens to know good sources, please add references! Anyway, until we have found better references, I am a bit reluctant to remove the Core citations, as we would have a large number of unreferenced redlinks (many of which are well-known conferences). Naturally we can remove bad conferences even if they happen to have a high Core ranking; just raise the issue on this talk page. — Miym (talk) 19:52, 3 June 2009 (UTC)[reply]
I think I found two more great conference quality assessment pages: ETH Computer Science Conference Search Tool, http://confsearch.ethz.chconfsearch and ArnetMiner from Tsinghua http://www.arnetminer.net/conference_rank/. They follow a scientific (really, they even have papers on it) approach using graph-based analysis of the citation graph. They also considers the libra and citeseer listing, garfield impact factor etc. confsearch is published and is being used by a prestigious CS group at a prestigious university. Of course they satisfy the criteria of not falsely pumping up bad conferences due to bias, as opposed to the highly unreliable Core ranking and the fake conferenceranking.org site ( http://www.rankingexpose.com/ ). They seem to misclasify a few excellent conferences, but this is ok.
Btw Miym, thanks for your great contributions to this article!
Thanks! — Miym (talk) 19:52, 3 June 2009 (UTC)[reply]
And please note that automatically generated rankings such as those provided by CiteSeerX are far from perfect as well. For example, [5] shows FOCS with ranking 266, right after "KiVS Kurzbeiträge und Workshop" and "ICINCO-SPSMC". Nevertheless, I agree that a high ranking on these lists can be used as a positive indicator of notability. — Miym (talk) 23:06, 2 June 2009 (UTC)[reply]
[edit]

Regarding external links: In general, before adding links, please read WP:EL and WP:NOTLINK. In particular, the policy says that "Long lists of links are not acceptable". There is a huge number of web sites with lists of events, and we cannot include all of them. Before adding a link, we should consider very carefully whether it provides a lot new information that cannot be integrated into the article and that is not yet covered by existing links (i.e., it is a "well-chosen link to a directory of websites or organizations").

An anonymous user has been trying to add the link http://events.informatics_events.org (which does not work, but should be http://events.informatics-europe.org/ instead, I believe). Currently that page lists only 31 conferences; moreover, these conferences do not seem to be chosen by using any particular criterion. The web site http://www.confsearch.org that we already have provides information about more than 2000 conferences. Both are comparable in the sense that they provide information about future events (including dates, submission deadlines, conference locations, and web site links). Therefore it seems that the link to http://events.informatics-europe.org/ does not contain any significant new information in comparison with the links that we already have; indeed, having several links makes it more difficult for the reader to find the most relevant links.

If you still think we should add the link, please explain your reasoning here. Moreover, please disclose any WP:Conflict of interest, i.e., if you are somehow affiliated with the organisations or web sites that maintain the lists. I am not affiliated with any of these, other than being a satisfied user of DBLP and confsearch. — Miym (talk) 07:16, 3 August 2009 (UTC)[reply]

Missing conferences

[edit]

In the field of quantum computation, I think QIP is the most prestigious conference. I'm not sure how these conferences are being sourced, so I'll leave it to somebody else to decide whether to add it and how to source it. --Robin (talk) 03:14, 17 December 2009 (UTC)[reply]

We used to have QIP on this list, and I tried to find any references that would show the notability of the conference but couldn't find any. I tried now again (checking both online sources and dead trees) and failed. Could you perhaps suggest some sources? In essence, any third-party source that explicitly shows that the conference is notable would be fine; typical examples are textbooks and survey articles that lists the most prominent conferences in a particular field, and conference rankings and impact factor lists. (I think QIP might be a bit tricky case: it seems that they don't have traditional published proceedings, and the submissions are just abstracts with 2-3 pages; some CS conference rankings might omit it simply because of these technical reason?) — Miym (talk) 16:18, 17 December 2009 (UTC)[reply]
Maybe you're right. I have no idea where to get a source establishing notability. I could find blog posts by major researchers which indicate that they attend the conference. But the attendance can just be seen by noting that the list of speakers at QIP usually includes all the big names in quantum computation, so the blog post probably doesn't help. --Robin (talk) 17:13, 17 December 2009 (UTC)[reply]
Another decent conference in QC is TQC. This is an actual conference with published proceedings and everything. Can you check your sources for this one? --Robin (talk) 19:08, 17 December 2009 (UTC)[reply]
Finding any sources for TQC seems to be a bit difficult for a different reason: it's a new conference; the first TQC was held in 2006, and they have had LNCS proceedings only since 2008. Many potential sources are simply too old to mention TQC. Usually this isn't a problem: conferences which are that new aren't usually notable yet; no-one knows whether such conferences will have long-term impact or not. I'm just wondering if we are too harsh to quantum computing, as the whole field is fairly new – but can we do anything else than wait a while and see if someone publishes new rankings that cover QC conferences? — Miym (talk) 22:38, 17 December 2009 (UTC)[reply]
Yeah, maybe let's just wait and see. --Robin (talk) 23:17, 17 December 2009 (UTC)[reply]

There isn't even a single robotics conference on this whole list. Are they considered to be mechanical engineering or something? ICRA regularly has 800 papers and ~2000 attendees each year. --WreckLoose (talk) 21:52, 15 August 2014 (UTC)[reply]

Grounds for inclusion

[edit]

Could someone explain what the criteria are for a conference to be included in this list? Should we include all the conferences in the top 50% of http://citeseer.ist.psu.edu/impact.html ? This website says "ICALP" is number 320, whereas Arnet Miner says it is number 7 for "algorithms and theory". A lot of references are to the CORE ratings, which are now inaccessible as they have now been taken offline as "obsolete". Should we remove the references to CORE? I'm a bit confused as to where this list is heading, and what it is for. ComputScientist (talk) 19:05, 17 December 2009 (UTC)[reply]

Criteria: Very roughly, the criterion is WP:N. I think the entries should be sufficiently notable to merit a stand-alone article. Each red link should come with references that show that if we created an article on the conference, it might plausibly survive an AfD. This is a fairly restrictive criterion, but we need an effective way to remove spam entries that this list seems to attract frequently, and we must maintain neutrality. In practice, this criterion might translate to listing at most 10 "best" conferences in each subfield. — Miym (talk) 19:55, 17 December 2009 (UTC)[reply]
I suspect that this criterion is going to lead to (has already led to?) situations where some of the conferences that are excluded in an area are better than some of the ones that are included. Despite that, I think it's a good idea to have some reasonably objective criterion such as this in place of subjective criteria that might be more accurate but are harder to verify. —David Eppstein (talk) 07:22, 18 December 2009 (UTC)[reply]
We can try to put together a list of some obvious omissions here on the talk page (see #Missing conferences above). Then whenever new ranking lists are published, we can cross-check them with these expert opinions. — Miym (talk) 10:28, 18 December 2009 (UTC)[reply]
Specific sources: I don't think there is any single source that we can use; as you have observed, various rankings seem to contradict each other. That's why I find entries with only one reference somewhat dubious, no matter which reference it is. So far the following seems to have been a fairly good rule of thumb: if there are multiple independent third-party sources, the conference is sufficiently good for inclusion here. I think quantity is in this case more important than quality: it doesn't really matter which sources those are (as long as they seem to be somewhat credible, e.g., textbooks from decent publishers, or rankings compiled by respectable scientists using somewhat objective methods). So, to answer your question: I don't think we should include anything simply because it is high-ranked in Arnet Miner, or in CORE rankings, or in an ancient Citeseer listing; but if you can find it in all of them, it might certainly be worth adding to this list. (The number of references can be also used to roughly order the entries according to their importance, or to prune lists if they get too long in some subfields, so in general, more sources is better.) — Miym (talk) 20:08, 17 December 2009 (UTC)[reply]
CORE: The list that I have used seem to be still available on the original web site: [6]; archived here. Thanks for noticing this, I'll fix the link in the article! — Miym (talk) 20:19, 17 December 2009 (UTC)[reply]
We have removed refs to CORE and replaced them with the updated ERA one. There are still many problems with ERA, but as mentioned above it is at least an indicator of notability. —Preceding unsigned comment added by John.legal (talkcontribs) 16:11, 27 January 2010 (UTC)[reply]
[edit]

http://social.microsoft.com/Forums/en-US/mas/thread/6d5704d1-ee39-45bd-86da-d1766a94344e

Until they resolve it, it is advisable to use sources other than http://academic.research.microsoft.com/ to justify inclusion and ranking. John.legal (talk) 10:20, 9 January 2011 (UTC)[reply]

They have resolved the issue and we can again use http://academic.research.microsoft.com/ for impact factor information. John.legal (talk) —Preceding undated comment added 11:31, 24 January 2011 (UTC).[reply]

[edit]

Hello fellow Wikipedians,

I have just modified one external link on List of computer science conferences. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

This message was posted before February 2018. After February 2018, "External links modified" talk page sections are no longer generated or monitored by InternetArchiveBot. No special action is required regarding these talk page notices, other than regular verification using the archive tool instructions below. Editors have permission to delete these "External links modified" talk page sections if they want to de-clutter talk pages, but see the RfC before doing mass systematic removals. This message is updated dynamically through the template {{source check}} (last update: 5 June 2024).

  • If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
  • If you found an error with any archives or the URLs themselves, you can fix them with this tool.

Cheers.—InternetArchiveBot (Report bug) 23:56, 20 May 2017 (UTC)[reply]

International Conference on Network and Information Systems for Computers

[edit]

Should this conference have a Wikipedia article?Oceanflynn (talk) 23:16, 22 October 2018 (UTC)[reply]