From jwl@compsci.bristol.ac.uk Thu Aug 27 15:20:02 1998 Return-Path: Date: Thu, 27 Aug 1998 15:18:40 +0100 (BST) From: John Lloyd Reply-To: John Lloyd Subject: Proposal for a major new conference series To: dannyd@cs.kuleuven.ac.be, klau@cs.man.ac.uk, pf@info.fundp.ac.be Cc: jwl@compsci.bristol.ac.uk MIME-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII Dear Kung-Kiu, Danny and Pierre, I'm writing to you in your capacity as leaders of the Program Analysis and Transformation field. A steering committee consisting of Joerg Siekmann, Chairman of CompulogNet, Krzysztof Apt, President of the Association of Logic Programming, and me as Chairman has been created to try to start a major new series of conferences on Computational Logic. This proposal therefore has the official support of both CompulogNet and the ALP. The background to this proposal is that after the Manchester JICSLP conference, I circulated a proposal to the executive committees of both CompulogNet and ALP to start a new conference series. My main motivation was the fact that the JICSLP series of conferences had been struggling for some time and after the Manchester meeting it seemed obvious that as a community we needed to take some action. On the other hand, one can make a good case that the Computational Logic community is as strong as it has ever been. It seemed therefore that a larger and broader conference was required to reflect this strength. (A more detailed motivation is given in the message appended at the end, which is a slightly edited version of the one originally distributed to executive committees.) At the same time, an even more important development was taking place. Joerg Siekmann had been negotiating with several other logic-oriented associations and, as a result, a decision was taken to form an International Federation on Computational Logic. I enclose a passage from a letter from Joerg and Krzysztof on this point. "Please feel free to mention in your discussions that a decision has been made on creating an International Federation on Computational Logic with Dana Scott as chairman, three or four vice chairmen drawn and eventually elected from the mayor subfields involved and who also represent the main geographical areas worldwide. The members of this umbrella organization will be the existing logic-oriented societies similar to the organisational principles of ECCAI. It is to become an umbrella organization over all existing logic associations and organizations, which would play a role similar to IFIP in computer science. This might make people aware that some irreversible changes are taking place in our field." Clearly the formation of this new Federation is a major development. In this context, the role of the CL conference series can be seen as integrating into a major conference series all the important aspects of the Logic Programming view of Computational Logic. (Of course, this doesn't at all preclude having other workshops or conferences on Logic Programming - the main point is there there should be one where as much as possible of what is going on in Logic Programming is on display.) The intention is that every four years the new Federation will host a mega-conference which collocates the major conferences in the various subfields of Computational Logic. The CL conference of that year would be Logic Programming contribution to that mega-conference. It is clear that the proposed new conference series cannot succeed without very wide support in the Logic Programming community. This is why I am writing to you as leaders of the Program Analysis and Transformation field. It would be very important for the success of CL2000 that the LOPSTR 2000 conference be included in CL2000. Therefore, I would like to ask you to approach whoever you see as influential in the LOPSTR community about the possibility of including LOPSTR in the CL2000 conference. Feel free to distribute this letter to these people. Please understand that at this stage the precise format of CL2000 is rather fluid and essentially all aspects of it are open to discussion. I would be happy to hear any suggestions which could make this idea work. I would be pleased to discuss this further with you by phone, if you wish. If you want to do this, please let me know your telephone number and a time when I can catch you. Alternatively, you can phone me at work on 0117 9545143 or at home on 01275 855329. I look forward to hearing from you soon. Best regards, John Lloyd ******************************************************************************* Proposal for a New Series of ---------------------------- International Conferences on Computational Logic ------------------------------------------------ Background ---------- This proposal comes out of some discussions I had with various people following the Manchester JICSLP conference. It has become clear (to me at least!) that something needs to be done to put some direction and focus back into computational logic (CL). This proposal suggests a way to do this. I believe a close examination of CL shows the field to actually be in very good shape. There are a number of very active subareas and some of these have excellent industrial connections. The major problem is fragmentation. We all gone off in different directions and now there is no event which brings us all back together again. I propose the way to address this problem is to start a (major!) annual conference series which would gather together the increasingly disjoint threads of CL and would provide an event which researchers in CL would be very strongly motivated to attend. I think there are good reasons for starting such a conference series: (a) It has all the advantages of the existing specialist conferences in the form of streams. (b) People are going to be motivated to attend a conference where, in addition to following their own specialist topic, they can get a broad overview of the entire subject from the top people. And they will want to publish in a conference series as major as this as well. (c) CL researchers are becoming so specialised that they are missing important ideas in companion areas which they should be exploiting. (I've seen this at first hand over the last couple of years.) The conference series can directly address this problem. (d) A major new conference series cannot fail to make a bigger impact than at present on ESPRIT (who fund much of our work), industry, and the CS community in general. The existing fragmentation is substantially diluting the impact we are having. We can change this - the CL conference series can become a prominent showcase for the contribution CL is making to information technology. (e) The symbolism of the start of a new century (and millennium) is much too good to ignore. A final point: I've been motivated to put this proposal together because this is the kind of conference I would be prepared to make any sacrifice in order to be able to attend! I strongly suspect many other people will feel the same way. Name of series -------------- International Conference on Computational Logic Short name: CL Frequency of conferences ------------------------ Annual (so the successive conferences would be named CL2000, CL2001, ...) Date, venue, and length of first conference ------------------------------------------- Late June, 2000. A conference centre in London. 5 days (Mon-Fri). Sponsors -------- CompulogNet, ALP, ESPRIT, EACSL, and others. It would be important to also be sponsored by some North American and Japanese associations. "Sponsorship" doesn't necessarily mean contributing some funding, but clearly that would be best. However, it does mean giving the conference official moral and practical support. Attendance ---------- 400 minimum. Proceedings ----------- A single proceedings for the whole conference published by a major publisher (e.g. Springer, MIT Press, ...). I estimate its length at around 1300 pages (using A4 pages, double column format, 8 pages per paper). This can be published in 2 volumes. Registration ------------ A single registration fee covers everything. Main ingredients ---------------- 1 keynote speaker 7 invited speakers 12 advanced tutorials (in 3 parallel sessions) 10 area streams (13 possible ones are listed below) 150 submitted papers for the streams (in 5 parallel sessions) 10-20 papers per stream A 2 hour panel to end the conference First 4 mornings taken up with invited talks and advanced tutorials Afternoons of first 4 days and most of last day taken up with submitted papers (with 1/2 hour slots for each paper) Industrial exhibition (in parallel) "Included" and collocated conferences (see below how this would work) Organisation ------------ Conference Chairman (i.e. local organiser) and committee Programme Chairman, plus small committee, who has overall responsibility for coordinating the technical programme (inc. choosing invited speakers) Stream Programme Chairmen plus programme committee for each of the streams who select the papers for their respective streams Likely streams -------------- Agents Automated deduction systems Constraint programming Database and knowledge base systems Implementation and architecture Knowledge representation and reasoning (inc. NMR) Machine learning Natural language processing Program analysis and transformation Programming environments Programming languages and systems Software engineering (inc. formal methods) Theory (this covers essentially topics of CSL and LICS) Notes ----- 1. The existing conferences which would overlap in some way with the CL series include JICSLP, CP, LOPSTR, PLILP, SAS, PEPM, CSL, LICS, IFL, POPL, ILP, CADE, ALP, PAP, PADL, and ICFP. We should declare the CL series to be the "continuation" of the JICSLP series (which would effectively end). Some of the above conferences would probably choose to be "included" in CL2000 (e.g. LOPSTR and PLILP). This will be possible (indeed, strongly encouraged) with the following understanding: (a) The stream titles are primary; an "included" conference takes places *inside* some stream. (b) "Included" conferences do not have their own proceedings. (c) Uniform standards of refereeing and acceptance will be applied for all streams. In any case, the programme committees for each stream would be similar in size and constitution to conference committees as we have now for the medium-sized conferences and would have considerable independence, subject to condition (c) above. Collocation will also be possible. This has the usual meaning of simply holding the collocated conference at the same place and time as CL2000. Collocated conferences would be fully responsible for forming their own programme committees, technical programmes, conference proceedings, etc. It would be best though to try to arrange special deals for registration at both conferences and also have such things as shared invited speakers. 2. The keynote speaker, who will open the conference, needs to be the "David Hilbert" of CL. We need someone who has a very broad understanding and vision of CL who can set the scene by giving the major technical and other issues we will have to address over the next few decades. 3. The invited speakers should be the most influential researchers in CL who can also provide a glimpse into the future. All invited speakers will be required to have full papers in the proceedings. 4. Advanced tutorials should provide an introduction to the subject matter of the streams suitable for someone who is an expert in CL but not necessarily in that stream. 5. The panel at the end will be concerned with a topic of broad general interest to people in CL and should attempt to summarise and draw lessons from what has gone on during the conference. 6. I've sketched out a plan for the 5 days into which all the events fit, which confirms the viability of the proposal. The venue will require a main hall to fit at least 500 (even if we only have 400 we need empty seats; otherwise, it becomes very complicated trying to find somewhere to sit) and 4 lecture rooms to fit between 100 and 200 people. Space for the industrial exhibition will also be needed. 7. The 2 year timescale is long enough for all the negotiations necessary with other associations and conferences. We should plan to announce the conference by the beginning of 1999. This gives people 12 months to prepare papers. The paper submission deadline would need to be at the end of 1999. 8. The location of London would be ideal: it has the easiest access for international travellers, the local language is English, it has good conference facilities, and it's an attractive venue especially in June. 9. I've given a lot of detail to make the proposal as convincing as possible. Of course, every detail is open to discussion.