5/09/2013

A note on bogus/fake conferences

"Change log": The post remains unchanged as released 5/9/13 9:54 PM, Central European Summer Time, modulo fixing typos, and summarizing the responses from WORLDCOMP; see past the disclaimer below.

In this post, I am concerned with the topic of bogus/fake conferences which has received attention specifically across the computer science community. If necessary, run this Google search to get some data points.

Disclaimer: Please, apply personal judgement when drawing conclusions. Avoid premature judgement; take into account that cyber harassment could be behind some of the information you find online.

Actually, I am not in the business of discovering bogus/fake conferences or even working out a solid definition of that notion, but I am affected transitively, as I discuss below. Also,  I may have an opinion about some related phenomena. For instance, IMHO, the notion of fake/bogus conference, as I see it presented online, is too much of a Boolean domain where I think instead that multiple dimensions are actually involved, each one deserving a discrete or continuos scale well beyond information capacity of Booleans. For instance, I have received (and perhaps even submitted) embarrassing reviews for rank B conferences. What do we call those conferences?

In the following, I tell you two non-fiction stories. The first story explains how I published a paper IMHO sub-optimally at PDPTA 2011 (part of WORLDCOMP). The second story details how the "John Peter" dynasty sends threatening emails to me because of some of my tweets related to WORLDCOMP.


BEGIN DISCLAIMER

I am not saying or implying, in any form, that I (Ralf Lämmel) classify WORLDCOMP as fake/bogus conference. In fact, I am grateful for having a paper published in the PDPTA track of WORLDCOMP 2011 and I hope WORLDCOMP can get further beyond all accusations and it can recover further from the alleged cyber attack, which I condemn. Also, I am not accusing WORLDCOMP or its authorized representatives to have sent the messages reported below.  

I do expect a confirmation from WORLDCOMP that indeed they did not authorize those messages. I have contacted WORLDCOMP as follows.

From: Ralf Laemmel
Date: Thu, May 9, 2013 at 10:02 PM
Subject: Fwd: Defamation of WORLDCOMP
To: sc@world-comp.org
Dear Sir or Madam,
I have received the following email from John Peter, who appears to represent WORLDCOMP. I would like to kindly ask whether WORLDCOMP has authorized this sort of email? I consider the form of this email as an attempt to threaten me.
Thank you and with kind regards.
Prof. Dr. Ralf Lämmel
Software Languages Team / Arbeitsgruppe Softwaresprachen
Universität Koblenz-Landau

END DISCLAIMER

SUMMARY of WORLDCOMP response

I had received a response and follow up emails within less than 2 hours after the initial post and my simultaneous email to sc@world-comp.org. I quote from the first response: "Once again, worldcomp would never send emails to any individual or listserv from email addresses other than @world-comp.org or amgsolo@mavericktechnologies.us. [...] The person who sent you the message below is [sic] to aggravate you (with the sole intention of creating enemies for worldcomp)."

More details and arguments were provided to me and based on my current understanding and careful judgement I assume that WORLDCOMP is not involved in the threats discussed below, but rather that the threats are part of the cyber harassment of which WORLDCOMP and some of its representatives have been the victim. In fact, I am looking at documents on a (partly) confidential basis which I would judge as substantiating a very substantial cyber attack that is hard to comprehend. This is even worse than I thought.

5/10/13 0:05 PM, Central European Summer Time

END SUMMARY

The paper "MapReduce with Deltas" at PDPTA 2011



Not listed on DBLP!

As it happens, DBLP has stopped listing (as of writing this blog post) PDPTA events. The last edition listed is PDPTA 2010. Thus, our paper - of which I think highly - is not listed on DBLP, which I find annoying because DBLP is such a useful tool in Computer Science. If a paper is neither listed on DBLP, nor ACM DL, nor IEEE explore, then the paper is much less discoverable and some may consider the paper unworthy of reading and citation. Of course, there are exceptions where great insight or strong results have been published in different ways. 

Disclaimer: Don't read anything between the lines. I am not saying that conferences that are no longer listed on DBLP are potentially fake conferences. Neither am I saying that conferences or events that are continuously listed on DBLP are likely rank A or B (or C or what have you). Neither of these two statements would make much sense, if you understand how DBLP works. 

I am just saying that PDPTA's listing on DBLP was a necessary condition for me to even consider the conference. Now, without DBLP listing (or any other helpful listing IMHO), I consider this paper to be published in a suboptimal way, just on the grounds of indexing. 

Why did I publish with PDPTA?

In early 2011, a research project with my student David Saile came to an end and we were looking for a suitable venue for publication. The project was concerned with a technical and conceptual innovation regarding the MapReduce model for data-parallel programming. Just for the record, data parallelism is a research area in which I consider myself relatively knowledgeable as I would like to substantiate on the grounds of the journal paper "Google's MapReduce Programming Model -- Revisited" which you might know if you are doing research on MapReduce and friends. 

As I was looking around for conference deadlines, I spotted the PDPTA track at WORLDCOMP. Of course, I also spotted several more deadlines of other conferences that seemed to have data parallelism / big data explicitly listed. Quite obviously, we could also have submitted to more generic and well-known venues concerned with programming and data in general. However, we wanted to get in touch with more people interested specifically in parallel distributed programming and big data and the sliding deadline of PDPTA ended up being perfectly aligned with our timeline. So PDPTA made it.

As a more senior researcher, I think quite about where to submit. Some degree of conservatism is involved. (I am not conservative on the scale I observe among my peers.) I want my papers to be exposed to the relevant community and there should be reasonable chances for other researchers to follow up on my work. These are dimensions of personal judgement relying on experience. Clearly, I also prefer better ranked conferences over the others. I am doing research in relatively different areas. So for me, it is natural to look around for venues that I don't know yet.

Back then, I didn't think of our results and writeup as being ready for submission to a rank A conference, also in the view of the need to publish soon, as my student was leaving university eventually. So I didn't consider venues such as ICFPVLDB, or others with the strongest track record. I sensed that PDPTA promised impact well in line with the results and the writeup at hand. Add to this that WORLDCOMP made strong claims about citation-related ranking and selective peer review. So I was hoping that PDPTA would count as rank B conference (eventually). Very briefly, I looked over papers from the previous edition just to see whether they are relevant in terms of topics and format. This looked good. (It happens that we did not reference any prior PDPTA papers.)

There were signs that made me wonder.

Perhaps. a repeatedly sliding deadline is something to feel suspicious about, some might say. Again, I am not very conservative.

Perhaps, the lack of finding relevant, prior PDPTA papers is something to feel suspicious about, some might say. However, during our related work study, we did not go specifically through prior PDPTA editions which, perhaps, we could have done. Rather we applied keyword-based search to literature indexes (DBLP and ACM DL) plus chasing related work from the references of key papers, as usual. We could have hoped for reviews to point out more related work, specifically also from prior PDPTA editions.

I wouldn't recognize anyone on the committee page. That's weird! If you pointed me to pretty much any international conference on programming or software engineering, I would look at the program committee and I would spot some colleagues, co-authors, well-known community members, or some researchers with whom I have worked in some program committee over the years. Anyhow, pragmatism kicked in and I decided to give PDPTA a try.

I would run though a simple Bing/Google search to see whether WORLDCOMP pops up in the context of bogus and fake conferences. SCIgen was an eye opener back then and we are more careful ever since. I would really hate to spend budget and brains on a bogus/fake conference. One can find much online information about bogus/fake conference these days. Even before the online trend to "leak" bogus/fake conferences, I would naturally disregard some conferences as they looked  clearly so unfocused, tourism-oriented, and "hypoglycemic" scientifically.

PDPTA looked pretty focused and scientific. I remember clearly that WORLDCOMP did not show up in any obvious way back then when I was running the search. I do remember finding one critical comment by someone apparently dedicated to the cause of revealing bogus conferences. That comment essentially said "The conference is not classified as a bogus conference as of now, but there may be some suspicious signs so that the conference is on the watch list."

Acceptance aftermath

Just to be clear, I didn't receive reviews for my paper (see the notification above). My student went to the conference and was modestly excited, but then again, it was his first conference. Also, PDPTA may just have attracted great papers over the years! Who knows! With big conferences that combine many tracks, it wouldn't be surprising to hear that quality varies across the tracks. I have heard such statements from people who I trust for the HICSS conference series as well as the SAC conference series. At this point, I don't know anyone senior first hand who is telling me anything substantial about WORLDCOMP. 

I want to believe that PDPTA was a great conference over all the years. 

Some transparency would help here. Anyway, not receiving expert reviews makes me feel uneasy. I consider peer review essential to how CS conferences currently work. I am well aware of the continuos discussion on this matter, as to whether peer review is the only way to go and how exactly peer review should be exercised. PDPTA promised and claimed peer review and selective acceptance. I am not completely convinced from what the evidence that I have seen. Again, some transparency would help here. 

Just for the record, I spotted a citation of our paper on ACM DL, but it is not counted on my bibliometrics page at ACM DL presumably because there is no standardized entry for the paper (such as those coming from Springer, ACM, IEEE, and other publishers and data integration facilities). There is no reliable way to search for other citations, but that's sort of a problem with ACM DL, which works best as an indexing service (not surprisingly) for ACM publications and works well for some other sources but not necessarily for all established ones. If our paper was at least indexed on DBLP, but it isn't.

Anyway, my student and me spent considerable effort on the paper and the underlying research and seeing the paper insufficiently indexed makes me feel sad.

The "John Peter" dynasty

Just hypothetically, suppose, you are suspected of running a bogus/fake conference and there are people (say like me) who spread the word about related suspicions. What would you do? This is not an easy question, certainly not for me, as I have zero experiences with being suspected of running bogus/fake conferences. However, I think I would NOT threaten people who express or share suspicions because I can't see how I would hope to contain the problem possibly this way. 

I would go for transparency and trust. That is, I would look at the claims and comment on them one by one, in detail, and with evidence at hands. For instance, if researchers suspect that the review process was compromised, then I would produce evidence and witnesses to show otherwise. If I cannot do that, perhaps because the claim is true, then I have to face the consequences. Perhaps, if some claims are true and others aren't, then the conference can be saved and the reputation of those involved can be protected, but without transparency all is at risk. 

If you want me to reconsider a statement I made, be it in a tweet or otherwise, please don't threaten me anonymously, rather contact me in a sensible way. Thank you!

Therefore, I want to believe that the following messages from the "John Peter dynasty" are not authorized by WORLDCOMP. Rather, I can imagine that indeed some complex form of cyber harassment is at play here, as possibly suggested by other sources like this.

As evidenced by a discussion on ResearchGate, I am not the only victim of such threatening.

Exchange with John Peter I

(If you don't know Hushmail - it's "Email with privacy".)

Here is an Email from John Peter I.

From:
Date: Sat, May 4, 2013 at 6:18 PM
Subject: Demand to remove your defamatory posting
To: rlaemmel@acm.org
Hi
We are upset to see your posting at https://twitter.com/reallynotabba/status/302889284724224001  about our WORLDCOMP. Your statements are clearly defamatory and we formally ask you to remove this posting within 24 hours from now. Else, we will follow legal actions against you and your institute. Consider this message highly time sensitive.
Sincerely
John

Here is my reply:

From: Ralf Laemmel
Date: Sun, May 5, 2013 at 12:16 AM
Subject: Re: Demand to remove your defamatory posting
To: johnpeter1@hushmail.com
Cc: rlaemmel@acm.org
Hi John,
Sorry if my tweet upset you,
I removed it,
Regards
Ralf

John refers to the following 91 days old tweet: "OMG, #worldcomp is actually a fake conference! A little suspicious it was, but some tracks looked quite good. worldcomp-fake-bogus.blogspot.de". In retrospect, I am indeed not proud of myself for choosing one specific link related to claims about WORLDCOMP because it may be interpreted as if I would have been able to check, to some reasonable extent, the specific claims on the specific linked blogging site. Further, I quickly realized that elements of cyber harassment, were also at play, given all the online pages and the tone used on some of the pages, which I indicated in another tweet shortly after 302889284724224001. So far, John Peter did not complain about the tweet considering cyber harassment. 

Back then, when I shared suspicions through the tweet, I made an effort to search for transparent detailed responses to the accusations some of which would be compatible with my personal experience. I am still looking for the ultimate responses. I would very much prefer an outcome such that the value of my PDPTA publication is reinstated.

No word from John Peter II

Did anyone receive any email from John Peter II?

Exchange with John Peter III

Here is an Email from John Peter III.

From:
Date: Wed, May 8, 2013 at 4:33 PM
Subject: Defamation of WORLDCOMP
To: softlang@uni-koblenz.de
Hello,
Thanks for removing your twitter posting on WORLDCOMP but you came up with another strategy to defame WORLDCOMP. Do not act too smart. I ask you to remove “WORLDCOMP” and any reference to it from all your twitter and other postings. We will directly contact your university officials and will also file a lawsuit unless you comply with our demand within 24 hours from now.
Sincerely
John Peter

Here is my reply:

From: Ralf Laemmel
Date: Thu, May 9, 2013 at 4:56 PM
Subject: Re: Defamation of WORLDCOMP
To: johnpeter3@hushmail.com
Done.
Ralf

The mail bounced, though. John, hello, can you hear me?

This is about the following tweet and associated communication; I have grayed out responses from another twitter user who tried helping me with my effort.



Lessons learned


  • Submit to conferences that you know. ("You" as in yourself or your supervisor, your experienced co-author, etc.)
  • If you want to step on new ground, check whether the targeted conference has a critical mass of people involved who you can trust (personally) that they hold up the scientific quality of the conference.
  • When a conference is labeled fake/bogus in a yet to be defined, affirmative sense, then we should allow people to resubmit their work for peer review, if they published their work with the conference in question just before the news broke, also assuming that it was insufficiently obvious, also in need of a definition or decision by a suitable board.
  • Do not close the door until suitable responses and evidence have been judged properly. Consider the possibility and the role of cyber harassment.

Prof. Dr. Ralf Lämmel
Professor of Computer Science
University of Koblenz-Landau



2 comments:

  1. Just a short explanation on what happened yesterday (as far as I know). [This is some kind of executive summary for Ralf's blog note]. Ralf will correct me if my understanding of the facts is wrong or incomplete. Of course this is only a summary.

    1) Ralf and a student sent a paper to the WorldComp conference. Paper accepted with no reviews.
    2) Ralf raises on Twitter some legitimate questions on the level of this conference and particularly on the absence of reviews.
    3) Some people (including me) retweet this information and try to help locating WorlComp reviewers to check if some testimony may invalidate the first (bad) impression.
    4) These people receive on their personal mails threats of legal action and pressure on their administrative hierarchy if they don't withdraw immediately their twits. The subject line is: "Defamation of WorldComp". The origin of the mail is a person named John Peter (1, 2 or 3 according to the time) and using hushmail. Apparently this is a campaign coming from WorldComp to defend the level of the conference and intimidate people who ask questions.
    5) After some episodes (...) it turns out that this person John Peter is not trying to help the WorldComp conference, but on the contrary to put them in trouble. Ralf is not the first one in this case.
    6) The WorldComp administration contacts us to dissociate themselves from the initiative of this person John Peter. They provide evidence of some previous cyber stalking. They claim that this person is their enemy and may change e-mail every day. His goal is to harm them.
    7) After looking at the evidence, and until further development one is tempted to believe the administration of the Worldcomp conference. This says nothing about the scientific management of papers at this conference, but apparently they are not involved in the Twitter censorship business which is good news.
    8) This was the situation yesterday very late. Hope this is the final state.

    There are some lessons that may be learnt from this.
    1) Better to avoid submitting papers to unknown conferences with PCs and organization committees that you don't know since some time and really trust.
    2) Be extremely cautious when you get a paper accepted without sufficient reviewing data.
    3) It's a good idea for conferences to have an official presence in the social media like Twitter. When forgetting this, conference organizers may open the door to uncontrolled communication. The idea of a "social media chair" in classical conferences is becoming more and more important.
    4) Never be impressed by guys who want to enforce censorship on social media like Twitter by bullying you. They may threaten you of legal action or of "contacting directly your academic administration". The good choice usually is to try to identify the source of the threats which often eliminates anonymous pressures. If the source is a real person, discussion works usually better than confrontation.

    ReplyDelete
  2. Thank you very much, Jean, for doing this summary. This adds an insightful view on the situation, certainly one I find very reasonable morally and operationally. We don't know nothing for sure, but I am very much annoyed by the cyber stalking at work. That needs to be stopped.

    ReplyDelete