SIG Governing Board Meeting
Friday, October 5, 2012
1.0 Welcome (Altman)
1.1 Welcome, Introductions (Altman, Madden)
The microphone was passed around for those in attendance to introduce themselves.
2.0 Report from ACM CEO (White):
John White introduced himself and thanked everyone for giving him a chance to provide an update on ACM. He explained that he and Joe Konstan will discuss some changes within Publications. An update on the state of ACM as well as the organizations’ priorities follows:
Membership: ACM has 104,000 members (as of August, 2012) and over 700 chapters around the world. This number has increased through efforts made with other organizations like CCF; ACM Membership is now included as part of CCF membership. Over the last 10 years ACM membership has seen a shift with most new members being outside of the US making ACM an international organization.
Finances: FY ’12 was a solid year on the general and SIG side. ACM is sound and extremely healthy.
International Initiatives: We are continuing our initiatives in India, China and Europe. We have been working with the Brazilian Computer Society and others in South America.
International Initiatives in India: We continue to work with the India Council to enrich the Computer Science community there. ACM India is the first ACM region to elect a council. Our education initiatives are moving forward with Ed Board involvement. There is a new group in India similar to ACM-W which holds a meeting similar to Hopper. There are many SIG-like groups building on the national SIG model. We are moving forward with India-wide SIGs, using the regional chapter model.
International Initiatives in China: We continue to do a lot of work with CCF. CCF has begun to raise dues for all renewing members to cover ACM membership. We have been sponsoring Turing winners to appear at the CCF annual conference (CNCC). We translate a section of the CACM for Regional Members in China. We are continuing to work on launching more ACM chapters. We encourage more ACM SIGs to hold conferences in China.
International Initiatives in Europe: ACM Europe has been active not so much with membership but in the way of visibility. Projects like “Informatics Europe” is a CRA-like initiative based in Europe. The Ed Council is looking at a project similar to the US report “Running on Empty” as there is room and a role for an Ed Council presence in Europe. SIGCSE is looking at participating. There will be ongoing work looking at establishing formal CS Ed standards there and might involve working with the EU. Plans are underway for the first ACM Europe Computing Research Congress (ECRC) which will be held in conjunction with CHI ‘13 in Paris May 2-4, 2013. We held a successful ACM Europe Chapters workshop. We continue to hold many ACM conferences in Europe.
Education Initiatives: Most initiatives are run by the ACM Ed Board and Ed Council and focus around CS 2013 and other major, multi-year events such as the 10K Project which is working to revise CS Ed standards with the help of 10,000 High Schools and 10,000 teachers. ACM is increasingly involved and is co-leading this with the NSF. We are also involved in the TauRus project which is surveying non-PhD institutions to determine what Computer Science courses they offer. We continue to work to make these committees strong, keeping in mind the necessity of improving Computer Science Education.
Turing Celebration: This event has been regarded as a success. We had over 1,000 attendees, and sponsored 70 students from around the world. The SIGs contributed – it was an all-around great event – we have been told we should continue to have more forums with Turing winners in the future.
Heidelberg Laureate Forum: This is patterned after the Landau award which alternates awards between Computer Science and Mathematics. We are just beginning to contribute to this and are setting up a sub-committee to establish ACM’s role. The goal is to decide participants.
In summary: We need to see more SIG activity outside of the US. I encourage you to develop conferences in Europe, contribute to the 2013 ECRC conference and engage interest within your community for the Heidelberg Forum.
3.0 Publications Report (John White and Joe Konstan):
White reviewed the proposed changes to the ACM publishing model. See attached slides.
White provided some background: ACM has been a scholarly publisher for nearly 60 years with publications being a significant and defining part of the organization. It is defining in the sense that ACM has a reputation, well earned, in publishing high quality content that covers a broad swath of computing literature. As much as that is the case the DL has become a defining part of the ACM Publications Department. It is the heart and the soul – the core of ACM publications.
As we look at possible changes at how we approach publishing, we look at our current philosophy. One of our goals is to support the community by providing high quality publications. The publications we have reflect the technical community. ACM has a history of granting significant rights to ACM authors. For over a decade we led the way by allowing authors to archive their papers on individual, institutional or funding agencies mandated archives, and then a couple of years ago we introduced AuthorIzer links which allows ACM authors to provide free access to the definitive versions of their papers. Another part of the philosophy has been to provide a real, state of the art, Digital library – not just collection of PDFs – but a digital library richly interconnected with lots of great features. Another part of the goal and priorities has been to price ACM publications and the DL incredibly affordably. So, in summary, the goal has been to be a fair access publisher in the digital age.
The DL and thus all of ACM content is available in thousands of Universities and Corporations worldwide. There are over 1,250,000 regular users of the ACM DL. Almost all researchers have access to it through their institutions. Almost all ACM content is freely available via self-archiving. So there have been some pretty dramatic results from this publishing philosophy centered on the Digital Library. The academic Librarian Community is extremely pleased with ACM both in terms of the content that we produce but also the features and functions of the digital library.
The pricing structure of the DL has ACM viewed as the good guys. There have been several positive results of the pricing structure of the digital library. While priced extremely low, the Digital Library revenue is now a significant part of ACM and a significant part of the SIGs.
White presented DL financial information. The DL revenue has become significant – both the gross revenue and the distribution of the amount that goes to the SIGs and the amount that goes to the ACM general.
This revenue is used to fund the fully loaded costs related to the Digital Library. That includes platforms, software license, development of features and functions, composition of all the content, metadata extraction, worldwide performance and Quality Assurance. It also supports the areas that are free to the world: bibliographic database inside of author pages, institutional pages, the work that goes on around citation linking and being able to compute bibliometrics, etc.
The DL revenue is so significant to the SIGs that it contributes in one way or another to the ability to subsidize membership originally done only through conference funds. It also funds ACMs obligation to advance computing as a science and profession and to seek computing in a socially relevant way – the “good works” of ACM.
The amount of DL revenue from 2012 is 16.3 Million. The fully loaded expense of the DL is 12 million dollars, which leaves a net of 4.3 million dollars. Based on the distribution formula; two thirds goes to the SIGs and one third goes to the general side of ACM. And that is not that different from the make-up of Conference, Proceedings and Articles and Journal Articles in the DL
There are a lot of things that go into the expenses in order to support a high quality collection like the Digital Library. This includes metadata, tagging with the new computing classification system, being a part of the publishing community and fully engaging in the DOI world. Running systems that keep the DL in a state of perpetuity that meets the needs and concerns of Librarians, enabling a locked, clocked and porticoed status, composition of the content that goes into the DL, editorial support, manuscript central, worldwide support, software development, data center, statistical packages, marketing promotion, member services and finance (to name only part of what is involved).
A group discussion followed on making the expenses more transparent and accessible by providing detail on the costs of the DL. There was a question as to why ACM membership does not include DL access. White explained that subscribers make up the 2.3 million in revenue. We will not be able to contemplate giving DL membership away to ACM members until a suitable replacement for this revenue has been found.
White indicated that there has been a ground swell surrounding open access issues over the last 12 months and the concern ACM has heard is that subscribed content is costly, locked up and unavailable. The idea that the subscription model is commercially exploitative is only weakly applicable to the kind of scholarly publishing ACM does but ACM is caught up in the discussion regardless. There have been increasing numbers of government mandates that research supported by different funding agencies should be publicly available. University policies are in the same space. There are boycotts of commercial publishers. There is lots of discussion within many of the SIG communities about open access.
This presents a bit of a dilemma for ACM. The Librarians, corporations and individual subscribers are generally happy. The content is high quality and it’s a good price. We are generally viewed as good guys not the bad guys. But some are unhappy. Some SIGs are unhappy and some SIG members are unhappy. On one hand this is about managing perception. It is a problem if ACM is seen as impeding progress. How can we accommodate the diversity in our constituent community, sustain a publishing program and still be viewed as and at least be out front at some level. We have been looking into this over the last 6 months and have developed a recommended set of changes.
The changes represent work from the Publications Board, the SIGs, ACM HQ and discussions with other publishers. There is a whole series of one-on-one interviews with leading researchers. SIGs like SIGOPS, SIGPLAN and SIGMOD and others have contributed. ACM Library Advisory Board also provided feedback. We learned a lot doing this. Others are fine with the way ACM operates. This is true across SIGs and within SIGs. There is a lot of diversity of opinion. We learned that Librarians will stop subscribing to the DL if our content is freely available in some aggregated form.
The Business model for the DL should ensure the long-term viability of the DL and be committed to pricing publications in ways that are fair and equitable. ACM can and should sustain its program by being the sole source of the aggregated collection of its content.
A set of recommendations came out of all of the discussions with the ACM EC, ACM Council, and the ACM Pubs Board. The Pubs Board will have a final discussion on this next Wednesday, the ACM EC will discuss and make final recommendations on Thursday and ACM Council will make final decisions on Friday.
There are three key points behind the recommendations. 1) Publishing programs should contribute to the broader mission of ACM. 2) ACM can and should sustain its publishing program by being the sole source for the aggregated collection of full text content. 3) Within these two principles, we should be as flexible as we possibly can while having a coherent program.
There are three key changes which are being proposed by the Pubs board and looked at and discussed. 1) One is to replace ACM copyright transfer with a license. 2) Provide more options to open up conference proceedings, free to the world. 3) Offer an author pays option.
There are 2 other areas of recommendation. 1) Do more to support on campus institutional patrons use of the institutional reporting. 2) Implement an ACM DL research program.
White reviewed each of the recommendations. Replacing copyright with a license means copyright would remain with the author. The existing rights the authors have would be incorporated in the license (all the rights to sell, archive, to re-use and so on). Authors would still grant ACM an exclusive publishing right and authorize ACM to act on the authors behalf in terms of defending the copyright ACM would continue to handle plagiarism issues.
With regard to opening up conference proceedings, SIG membership already provides full access to all a SIG’s content in the DL; a SIG member benefit that has been available for a long time. SIGs will be able to opt to have the conference proceedings open in the DL before, during and after for a total of 1 month – making them freely downloadable. For co-sponsored conferences we’ll need special agreements to give access in this way. There are still some issues that the Pubs board and the EC haven’t looked at regarding co-sponsored events with other organizations.
Konstan indicated that if proceedings are to be open before the 1st day of the conference, it must be in the call for participation because authors have the right for their work to remain confidential until the date of publication - that comes up with intellectual property and other issues. This requires advance planning and is something you must decide at the time of the call for participation and solicitation of papers from your community.
SIGs will be able to opt to post a table of contents of the current volume of sponsored proceedings with authorizing links on either the SIG website or the conference website making proceedings freely available until the next conference or for 12 months. Essentially making the conference proceedings freely available through a table of contents and a set of AuthorIzor links.
The Pubs Board has indicated this will be a three year experiment which will be monitored so that we can compare free downloads versus subscribed downloads to gauge the impact on how the content is being used and if there is any impact on SIG membership. After three years we will either end it, adopt it or expand it.
A group discussion on the distinction between recommendations 2 and 3 followed. A clarification was made on the distinction between the AuthorIzor limits and the scope of option #3. With the AuthorIzer, as an author, you can have one Authorizer link. This option is independent of #3.
White continued with the third change to offer an author pays option for any article. It would mean there would be an author payment charge and upon payment the article would be permanently free. It will be covered with a creative commons license. We are still working on the right rates but there would be a lower rate for proceedings articles than there would be for journal articles and there would be a lower rate for ACM member authors versus non-member authors. We are looking at how to responsibly use this revenue, if there is any at to keep the subscription prices low or support the DL in a positive way.
A question was asked about extending the discount to SIG members as well. It was agreed that such an option would be considered.
Items also discussed: 1) improving access for individuals and institutions. 2) doing better job making the DL available for research purposes. 3) Aggressively communicating a highly visible and clear view of what is the ACM Publishing Model.
There was a comment about the unfairness of making people who may not have many resources pay as part of the “Author Pay” option. White indicated that was a concern but at the moment it is a solution that solves other challenges making it a worthwhile solution.
Konstan was asked to say a little about why the exclusive license to publish? He responded that there was a very simple answer, if you don’t have an exclusive license, somebody else can come and say, “oh we are going to have a copy of all of the ACM’s Publications in own collections” at which point libraries will no longer subscribe. If you don’t have the right to publish in its aggregated form, then someone else can copy it and present the aggregate copy but still allow individuals to make their papers published somewhere else.
There was continued discussion regarding exclusive vs. non-exclusive. There was concern that changing from copyright to exclusive license wasn’t really a change and the authors that are concerned about this will know that.
White indicated that the exclusive license issue was discussed a lot and many people had that opinion. ACM is not trying to say to the community, “we are going to a licensing agreement to pull the wool over your eyes”. We understand this is a particular kind of licensing agreement with an exclusive publishing right in it. We believe, at least right now, that to maintain the DL in an aggregated collection we need exclusive publishing rights. Beyond that, we are doing everything we can to let the community have free access to the stuff that it wants to read in a reasonable way. All of the things put together are important, whether it is licensing opening up conference proceedings or an author pays option.
Konstan added two things: the main difference between the license and the copyright transfer today, is that the license fixes the rights at the time of transfer. Copyright transfer, because copyright law can change and has changed, is something that a number of the librarians are advising some of the authors at their Universities against because what goes with copyright transfer may change over time and there was a lot of concern from some people over things like the code samples that are inside a paper and what does that do to the copyright of the code if the code’s in the paper? And that is why you notice the specific things that are going on about redistribution and open source and rights to code. The other way to view this, is that this is a huge and bold move. Saying that every conference we have, if the SIG wants it, the current year’s proceedings are out there for anyone who chooses to go get them, that is a big and risky step. The backup behind that is a step that is very conservative that says very few things are being done that can’t be studied when they are done and adjusted down the road.
The reason that this is a three year experiment is because we don’t know what will happen. The board debated what would happen if almost all the access came in through the free links, is that “my goodness we are hurting our revenue model!” or “my goodness, the open access people were right! We need to make these things open for the health of the community.” What if no access is through the free links? Is that saying that “look the free access people were wrong” or “there is no harm in indulging them?” We have no idea. We do want to do things based on data and that is the reason the plan is based on something we can track. Looking two years out, into this three year experiment, we need to decide whether to continue it before it’s over. We’ll be having discussions in about 2 years, looking at the total number of downloads and total number of accesses, where they are coming through, what is happening with SIG membership and what are we going to do next?
The whole issue of a conference proceeding being open for a year or not, is in the hands of the SIG leaders and it is fully expected that some SIGs will say that “we are going to make it a policy that every one of our conferences is open and other SIGSs will go the other way.” The other issues we are talking about are mostly significant but in incremental improvements. This is one that will affect SIGs substantially.
Konstan asked the SIG leaders to provide him with comments they’d like passed on to the Publications Board by Wednesday. He then gave other publication updates to the group. The new classification system is complete. Both categories have been mapped and will be seen in 2013. That is going to be rolled out to authors. He provided information on the Social Computing Journal. NSF, the CCC and the ACM are supporting a workshop as one of the outcomes of that. He mentioned the book program and asked anyone interested to contact him off-line. ACM is in the process of developing templates for single column conference proceedings. This is going to matter for two reasons. 1) as soon as we have some drafts, these are going to be drafts that in the word version at least are going to be features for accessibility we are trying to figure out what you can do in LaTex. I am going to be looking for SIG volunteers (people in your SIGs) who are going to provide feedback on the single column format. There is also a decision which we deferred and we are going to come back to you about. Do we really need two of them? Do we need a format for the archival works of the DL and a format for the other things that conferences publish?
Prior Publication: There is a proposal that has come out saying that the current practice of requiring a journal paper to be significantly extended or different from the conference paper is in conflict with certain communities with our SIGs and within Computer Science where often the only difference between a long conference paper and journal paper is how detailed the reviewing is. If you decide this is something we should spend discussion time on we will come back to it.
Copyright: We have said everything that should be said here. I think there is a take away message that I will take really strongly back that says this has to be labeled in a way that says clearly, “No, we are not pulling the wool over everyone’s eyes. There are other legitimate reasons for the change. Make it clear that this exactly what this is and that it is not substantially changing ACM’s rights. You have heard all that.
Pre-post: One of the reasons that the SIGs have opened their proceedings up during the conference is because they have given up on printed proceedings and they want a way that the conference attendees can get access to the proceedings at the conference. ACM has for many years made this available during the span of the conference. This is not new. This is being formalized to opening it up 2 weeks ahead because we get better discussion when people have read the papers before they come to the conference. Some people only read the papers after attending the conference. To do this requires planning. That is something you need to go work with your committees on.
Author pays: When you talk to the folks in the library field there is a lot of skepticism about publishers who say “we will do author pays as an option but we will still collect all of the subscription revenue.” ACM’s commitment here is part of what is making this hard. When we say we will reduce subscription costs to libraries if authors pay, what we are also saying is we have to have a solid business model behind this. That is why there is a lot of work in getting it right.
Conference pay option: This was discussed at the Pubs Board. We know conferences couldn’t afford this. The expectations of libraries are such that we need to provide a commitment to members. Since conference revenues are up and down, there are bound to be inconsistencies. Compared to Author Pay, this is far more inconsistent because with Author pay, you will never have an occasion when every author pays.
There was a question about the Hybrid option. Could a SIG pay for the best student paper award winners papers instead of giving them a prize? Konstan indicated that the Publications Board discussed that. Conceptually, there isn’t a problem with this but it will be discussed further. Konstan agreed to bring it up at the Publications Board.
A question was asked about what happens if a third party aggregator sponsored proceedings? Konstan responded that is one of the things the Pubs Board is concerned about. We are not ready for that. There is also the issue of competition between the SIGs that we didn’t want to get into. These are all good things to revisit in two years.
Konstan was asked if ACM had considered the model of counting the number of citations over x number of years? And then open it up if a paper gets more than a certain number of citations? Because it is a really popular paper and is considered seminal in a particular domain so it has become a free-for all reference. He indicated that the option wasn’t discussed. He asked to stay focused on the Author Pays option. This is not that the conference says we can choose that authors pay. This is not that you must pay to be published. This is anybody can be published, but if you want your paper to be free in the DL, beyond the other ways we have talked about, not your own personal archive copy, not through a SIG link, but if anyone wanders into the DL and come upon your paper it doesn’t check if they are a subscriber, then you pay an extra fee for that and for the right to open your paper through the creative commons copyright. We had discussions about whether we would have venues where everyone must pay and we rejected that at this time.
There are additional issues that are small but will matter to certain people. Individual subscribers complain that their institution has access but they don’t know how to use this remotely. We can now allow them to access the DL using the institutional password for a certain number of times. As long as they use the library regularly, they will never notice.
Konstan was asked if Google currently has access to Meta Data. He responded that they did and ACM will work with Google and others so that they can access the data well. That was the momentum behind the new CCS. Everyone says, who cares about keywords and indexing, Google takes care of that, but it is Google who cares about this because it helps them do their indexing.
Regarding the SIG Access Experiment for conferences: This decision can be made conference by conference so some of you have structures so that your steering committees have substantial authority over a conference. It cannot be made year by year. Very clearly the intent of the Pubs Board is that this experiment must be consistent for us to collect relevant data and not create a miserable experience for users. If you want the conference proceedings to be open for a year, if you are doing this for a series, it must be done for the three year experiment, we need to have the next year replace the previous year and not just have the thing disappear for some people who are complaining “why did you only have it for a year?”. If this decision is approved, you are going to have to start thinking about “do we want to participate?” Some of you clearly want to and you may want to set a policy that all of their conferences will do this, and it will be on the SIGs page under the big “open access” banner or “free access” banner. For other SIGs, this might be more complicated. Understand that when you open this up, you are under cutting one of the things you said you sold your members as a membership benefit. You are now telling prospective members: you get access to this …and so do non-members for a year. Make that decision intelligently. Have those discussions. Nothing happens so quickly that you can’t have the time to talk about this within your Executive Committees or Conference Steering committees.
Within the next year we are going to have to start thinking about how we are going to measure the success of the experiment. We are prepared to count downloads and changes in SIG membership but there may be things that are not obvious that we should be measuring and we need to know what those are.
It was suggested that ACM craft a different experiment focused on the communities which are currently the most vocal allowing non-exclusive access license. For some communities these changes are not enough.
Konstan summarized the comments from leaders: lots of support for experimentation and even for these specific experiments. A clear message that we should be careful not to make the change to the license seem like we are trying to fool people – make it really clear what the reasons for it are and what the limitations are. The idea that perhaps there should be a “refer-izer” link so a DL subscriber can mail a non-subscriber an authorized link to somebody. Discount author pay rates for SIG members specifically being put into the discussion. The fact that there is still even with these kinds of movements, discontent, can we still craft a non-exclusive license that would still work for the objectives that are there. Experiment in that way. There was some rebuttal to this point and a proposed option about limiting the size of aggregation. Can the SIGs pay an open access fee for Best Paper Award winner as a way of recognizing work?
Konstan emphasized that this is not a discussion that is ending.
4.0 SIGHPC (Pancake)
We are currently in transition status and would like to present our first elections next spring.
Our five-year plan for 2011-2012 laid out in 5 goals: Build membership quickly; focus on targeted, low-costactivities that will have almost immediate impact, to develop a sense of virtual community; work with ACM andSIGARCH to transition the sponsorship of SC to SIGHPC; work with industry contacts to identify activities they can fund that build awareness; position SIGHPC as spanning and disciplinary barriers to create new opportunities for collaboration.
Member benefits: Electronic newsletter, 3 year access to online information; member rates at conferences and workshops; access to HPC materials in ACM-DL; online interaction through social media venues.
Transition Plan for SC: SC has been sponsored by SIGARCH since its inception in 1988. Only tangentially relevant to SIGARCH. Only 3% of attendees are SIGARCH members. SIGARCH and SIGHPC both want SC transitioned as quickly as possible. Formal plan was put in place for the Spring of 2012. David Wood has been extremely helpful. SIGARCH continues to underwrite SC while SIGHPC builds up its fund balance. Both SIGS share responsibility during transition.
SIGHPC Finances: Intention is to build fund balance as quickly as possible and will largely be through future SC conference surplus. Five-year plan also said we would leave membership funds essentially untouched for firstelected officers. Only low-cost or “self-funded” activities. Balance at end of FY2012 was $17,503. Budget for FY2013 projects a balance of $86,328. Based on only 850 members, low-cost, immediate-impact activities.
Goal was to “develop a sense of virtual community”. Social media (John West). Followers: LinkedIn (739),Facebook (245), Twitter (219). Recognition program (Jack Dongarra); SIGHPC-initiated prize nomination; assisting several groups with other nominations; mentoring HPC scholars (Barbara Horner-Miller); arrangingvisits to non-academic HPC sites. Student gives seminar to peers afterward.
Industry-Sponsored Activities: Explored possibilities with EAPF. Industry requested we consider formalrelationship with group producing course materials. Some doubts about the quality of the yet-to-be-developed materials.
Will re-visit possible relationships once materials are available for review. We have a workforce development program. NITRD organization is developing strategic plan for HPC workforce development. Wereviewed/commented on drafts. Also are also helping them with state-of-the-practice survey.
Create Opportunities for: In-cooperation events. 3 to date (workshops in US, Europe, Asia). Working with ACM to expedite publisher relationships so major HPC conferences can be in-coop events. Two “incipient” chapters; 2 more are interested. We have a YouTube channel film called “Women of HPC”.
Recent additions: Members now can access SC and workshop proceedings in ACM-DL. . We also have bookgiveaways and have a “What we should be reading” column in newsletter.
Recommendation: The SGB approves the recommendation of the SGB EC and congratulates SIGHPC on their program performance and finds it viable to move out of initial transitional status. SIGHPC is granted full SIG status for the next 4 years.
5.0 Viability Reviews
5.1 Viability Review: SIGART (Gil)
Chair: Yolanda Gil, University of Southern California, Vice-Chair: Qiang Yang, Hong Kong University of Science and Technology, Treasurer: Gautam Biswas, Vanderbilt University.
We are considering changing our name to SIGAI. There was a show of hands in agreement.
Elections held in 2009: two new officers, one continuing. Chair: Yolanda Gil, University of Southern California, Vice-Chair: Qiang Yang, Hong Kong University of Science and Technology, Treasurer: Gautam Biswas, Vanderbilt University, Past Chair: Maria Gini (University of Minnesota)
Established an Advisory Board to establish good relations (includes current President of IJCAI and immediate past-president of AAAI): Tom Dietterich (Oregon State University), Jim Hendler (Rensselaer Polytechnic University), Haym Hirsh (Rutgers University), Eric Horovitz (Microsoft Research), Craig Knoblock (University of Southern California).
Established Liaisons for Educational Activities, Mehran Sahami (Stanford University), Peter Norvig (Google Labs).
Finances: We have a very healthy fund balance of $745K. Our DL revenue is stable. Membership is around 1,000 members. Retention for first year members is around 40%. Retention for second year members is around 80%.
We are discussing investments of funds that would allow SIGART to gain and retain members.
Member Benefits: Reduced registration at the conferences sponsored by SIGART. Conference proceedings from AAMAS, HRI, and IUI (all on CD-ROM). Reduced registration at many conferences held "in cooperation" with SIGART. Access to the following contents in the ACM Digital Library: Proceedings of all the SIGART sponsored conferences and some in-cooperation (doubled articles in 2010, maintained in 2011).
Community Benefits Provided by SIGART: Sponsorship and support of conferences relevant to the AI community; Autonomous Agents Research Award.;Travel support for students to SIGART sponsored conferences; Support for AAAI/SIGART Doctoral Consortium; SIGART Web site (14K unique visitors, 2K repeat visitors) (2.5K visits/month); SIGART Conferences.
Sponsored Conferences: ASE: Automated Software Engineering, September HRI: Human Robot Interaction, March IUI: International Conference on Intelligent User Interfaces, January K-CAP: Knowledge Capture, October (odd years) WI/IAT: Web Intelligence and Intelligent Agent Technology. We have a special agreement with AAMAS. We have many in-cooperation conferences with AAAI on Doctoral Consortium.
SIGART Goals I: Continue to support conferences, strong organization committees, many volunteers, high quality papers. Major issues this period: Negotiated a renewed agreement with AAMAS. We worked with SIGCHI to resolve IUI governance and continuity. We set up an Oversight Committee chaired by Alain Chesnais. We hope to continue to contribute to the Autonomous Agents Research Awards annually.
SIGART Goals II: Outreach to the community. We have a new web site with advertisement of conferences and other community events. An article promoting SIGART was published within the Chinese Computer Federation publication. Our student support involves AAAI/SIGART Doctorial Consortium. Travel grants for conference attendance amount to $25K a year. We support CS Education with our involvement with the IEEE/ACM CS2013 Curriculum Initiative. We hope to continue to work as Educational Activities Liaisons and are currently discussing new activities focused on recent graduates and early career researchers.
Recommendation: The SGB approves the recommendation of the SGB EC and congratulates SIGART on their program performance and finds it viable to continue its status for the next 4 years.
5.2 Viability Review: SIGecom (David Parkes)
The SIG is relatively new. It is the ecommerce SIG founded in1999 by Stuart Feldman. We are made up of about 250 members and our main activities are the EC Conference, Exchanges Newsletter which is new, and we have a new journal as well: ACM Transactions on Economics and Computation.
Our fund balance is quite healthy at around $210K. The main activity that we have as a conference is we are seeing generally increasing submissions and attendance. We consider this and I think it is generally considered the premier forum for research that is computer science centric in the area of economic commerce. We have high technical standards. We have submissions from a diverse community. We run a workshop program that has been well received and we had EC 2012 in Valencia and we will continue to have international conferences in the future as well. We have had co-locations in the past at a number of places.
We did get some press this year on NPR on an empirical study on the effect of Groupons on Yelp reviews.
With regard to finances, our conferences have been averaging a surplus of $19,000 a year since 2004.
There is a new journal, the ACM Transactions on Economics and Computation. We think this is a clear benefit that is important as it represents the first time a journal has been dedicated to this area. The editors in chief are Vince Conitzer from Duke and Preston McAfee from Google. We also have a Newsletter which is published multiple times a year and is doing quite well. This has a new editor, Arial Procaccia from CMU.
In terms of activities, we are formalizing our awards. We have 3 awards proposals which are under review by ACM at the moment. We do in-cooperation events we do supported events. For example there have been many workshops which have been happening recently and we have been supporting those.
For the Annual conference, we have been trying to embrace breadth. What we did for the first time is to have 1.5 tracks. We are also looking to co-locate with Decentralization in 2014.
One challenge is that our membership has fallen a little bit and I looked into this and it seems that generally around ACM about a third of the members are students but we have about 10% student members. I am interested in any feedback on how we can take that number up to the ACM norm. We are focusing on ensuring relevance and breadth as we go forward.
Recommendation: The SGB approves the recommendation of the SGB EC and congratulates SIGecom on their program performance and finds it viable to continue its status for the next 4 years.
5.3 Viability Review: SIGKDD (Ankur Teredesai)
I am Ankur Teredesai, I am the Information Officer for SIGKDD and I also have with me Samy Uthurusamy, the Awards Chair as well as a long term founding member of SIGKDD. We will are representing SIGKDD on behalf of Chair Usama Fayyad.
Our financial goals for the year have largely been to maintain a healthy growth rate and a balance of over 600K. Our current balance is about $1M. It has been consistently above that for 3 years mostly due to Industry contributions though they have been dropping we were very successful last year in China and we reached the record high level in sponsorships.
We continued NSF support for student travel awards. Usually KDD conferences end up with a surplus balance. We primarily support and outreach. That is a goal. We have seen an increase in student support to KDD conferences. We have also been successful in increasing the visibility around the Distinguished Dissertation Award as well as research & applications. We are committed to that. The membership benefits are pretty standard for our SIG.
We have several awards including best papers awards, best student papers, distinguished dissertation, best applications, innovation. We have service awards which are quite prominent and bring attention to our community. These are given to members who have been active in the community or Data Mining in general. Our Outreach efforts are important to us as is access to the DL.
Over the last year our membership has been over 1,200 which has been about a 5% drop from last year. The more worrisome factor for us has been the decline in student members. If any of your students are interested in Data Mining at all, please encourage them to give us $15 and become members.
Our goal #1 is to maintain the highest quality of research and application in terms of full publication and outreach. We are engaged in building an intellectual community and recognizing excellence within that community – both research and applied.
Our second goal is continued success with our flagship KDD Conference. Last year we had a record attendance of over 1200 people even though it was in China – or maybe because it was in China, we will find out next year.
With regard to sponsorship, we co-sponsor WSDM which is another popular Web Search and Data Mining conference and ICUC for Ubiquitous Information Management and Communication. We collaborate with SIGMOD and others on 6 conferences. The unique proposition for KDD has been the formation of the Industry Practice Expo. We initiated IPE as separate tracks within the SIGKDD conference. The idea of IPE is to get invited talks from top practitioners which have been a huge driver.
We fully sponsored the KDD Cup competition which has been a flagship contest within the SIG. Last year we had over 900 team entries in completion. It helps us recruit students for sustainability. We have continued to work on several engagement programs and special benefits. We also revamped web presence.
Recommendation: The SGB approves the recommendation of the SGB EC and congratulates SIGKDD on their program performance and finds it viable to continue its status for the next 4 years.
It was suggested that the leaders might want to make membership a requirement for students to receive travel grants as a way to boost student membership. It was also suggested that the ACM SIGKDD logo be placed on the website.
Unanimous approval of the recommendation by SGB
6.0 SGB EC Administrative Reports
6.1 SGB EC Update (Altman)
Erik Altman introduced the members of the EC, Elisa Bertino, Barbara Boucher-Owens, Yannis Ionnidis, Patrick Madden, David Pennick, David Wood, Vicki Hanson, Brent Hailpern, Joe Konstan, Andrew Sears, Donna Cappo.
In terms of ACM council, I think we have good representation both directly and indirectly from the SGB here. Cherri Pancake, Vicki Hanson, Alex Wolf are SIG members involved in ACM Council.
Erik gave a review of the SIGGRAPH Asia allocation. To mitigate loss and risk Koelnmesse now manages the conference. Last year SIGGRAPH received $77,000 from Koelnmesse and the EC worked together to determine how to assess an allocation for a program whose expenses fall outside standard practice. It was decided that a standard SIG allocation of 16% would be assessed from any revenue received
The ACM Distinguished Speakers Program is looking for SIG involvement and attendees are encouraged to reach out to Committee Chair, Gabby Silberman, or contact Erik.
Vint Cerf has encouraged the SIGs to develop listings of SIG members in the news in print or on-line communication as a way to increase the value of SIGs and increase SIG membership.
Erik updated the committee on the status of SIGHIT and explained that it was not taken out of transition because its programs did not meet expectations. As a result, the SIG is no longer operating. Joe Konstan will provide an update on the work that he and John White have been doing in the Health Informatics field and with AMIA. We are focusing on ramping up activity there.
6.2 SGB EC Task Force Updates (Madden and Wood)
The concern with nurturing new SIGs is that the infrastructure and the number of people that you need to have a viable SIG is fairly large. So there is a hurdle that we have to get over in order to propose a SIG.
David Wood reviewed the SIGNANOCOM Proposal: This Special Interest Group is on Nano and Molecular Computation. The proposed objective was to develop, foster and promote research towards a new highly interdisciplinary communication paradigm that enables communication and networking at Nano and molecular scale. The concerns that came up were the lack of an established community. The narrowness of the research idea there was a concern as well as how they were not encompassing work on Nano computation. An additional concern was the lack of established conferences other than an IEEE workshop.
Donna, Patrick and I had a conference call with all of the organizers involved where we expressed those concerns and the recommendation was that they broaden their scope to include Nano Computation and to start this new conference under the auspices of the SGB and to explore cooperation with the existing SIGs. I expressed to them that there is Nano Computation work being done at SIGARCH. There is also interest within the SIGARCH community. SIGARCH is interested in cooperating with them on a conference and along with other SIGs such as SIGCOMM. They are supposed to be reaching out to you and if you have interest you should reach out to them. Once this conference gets going we should gage the community interest for the new SIG and pretty much everyone agreed.
Donna Cappo reported that she received an email message from the NANOCOM folks and they are planning to send a proposal in the next couple of weeks.
6.3 ACM/AMIA Task Force Report (Konstan)
AMIA is the American Medical Informatics Association and, although they have significant operations outside of the Americas, as we do. It is an organization that is already involved in creating conferences within the space of Medical Informatics and Professional Opportunities and has been a partner in the discussion in how we can work together to advance the fields and address the challenges of Computer Scientists who have had challenges publishing with credit in AMIA venues.
Much of this is because they have a different conference ethos than we have in computing. I along with Patty Brennan, an AMIA appointee have been working together with committees from both organizations to put together a draft proposal about starting a shared conference series that would move around and attempt to engage different constituencies. Things are moving along in a pretty good way though not always fast.
The next step will be a conversation with our CEO and theirs along with other key volunteers. The goal is to have the conference steering committee chartered and going this fall for a conference which will happen sometime next fall. Once that happens we have a long list of collaborative activities that would help in areas of publications and developing research in the field. We will continue to update as things go along. At some future SGB meeting we will have an AMIA guest attend.
6.4 SIG Proposals (Wood)
There is nothing to add to the report on NANOCOM.
6.5 Professional Development Committee Report (Tracz)
I would like to make you aware of the no-cost opportunities available to you to provide some visibility to your SIGs and provide input to the general technical community.
We have created Tech Packs in most areas and there are a couple more in the areas of security and software assurance. We would love to have involvement in the identifying of authors and updating the packages. Completed Tech Packs: Cloud Computing, Parallel Computing, Enterprise Architecture, Mobility, and Business Intelligence.
A Tech Pack is an annotated bibliography referencing open source materials as well as materials in the DL. Learning Paths is self-explanatory, using these languages (Pyton, Ruby.)
With regard to the Webinar series, ACM is a little behind in the game. Right now we have had 5 seminars. The Webinars are very low cost and no cost to the SIG. Webinars to date: Cloud/SIGMOBILE, Recommender Systems/SIGCHI, Big Data/SIGKDD.
SIG Members have served as moderators and identified speakers. It is very easy to create one of these Webinars. They are VOIP, WebX type with charts only that get broadcast, get captured and then downloadable. In two weeks the Professional Development Community is meeting and we are hoping to have monthly webinars. Please refer any candidates who would like to participate to me.
We are hoping you can promote the work of the Professional Development Committee and that it is of use to your members.
It was mentioned that SIGGRAPH 2013 is looking to develop a SIGGRAPH University and they want to take some of our existing courses like Open DL and actually make it an open source course. This appears to be a good place to put something like this. Along with this I would like to tie-in annotated information.
Will Tracz: One of our activities is to create a “cookbook” on creating webinars and will involve shipping A/V equipment to your conferences or events so that you can capture keynote speakers or panels and once again, after post production but them in the Learning Center or link to the website.
Continued Discussion on Publications (Joe Konstan):
There has been some expressed concern about the issues that will be on the agenda next Wednesday regarding the subsequent publications of conference papers. The concern centers around the issue of the long standing ACM policy that says if a work has been previously published, it cannot go into an ACM journal without substantial addition of new work. There are differences within the community with regard to the review processes for conference work and journal work.
The group reviewed in general terms the number of errors identified in both review processes. There is a loose percentage of about 80% from both, with high variability on time limits and multiple iterations within both. This issue has been a concern with SIGCOMM with the method of disclosure being of concern. A comment was made that there is a need for clarity on what it means for a paper to been labeled “the Journal version”. There have been anecdotal comments about the ease of refereeing a paper is greater than the ability to finding time to review it. The suggestion was made that a disclaimer of sorts might be a solution. The answer might be to have different labels other than “referee and review”.
7.0 Best Practices Session
All SIG attendees where dived into 5 groups which broke apart to discuss pre-selected topics.
Groups 1 and 2 discussed these questions and chose a representative to deliver their thoughts to the larger group:
- Does your SIG routinely extend officer terms?
- If so, why?
- How much of a challenge is it for you to put together a slate with multiple candidates for each office?
- How long a term is required to complete major initiatives?
- How long does it take to learn the ropes and become an effective SIG Chair?
- Which do you prefer: (a) 3-year cycles with mandatory elections or (b) 2-year election cycles with officer-reappointment allowed?
- Does your SIG have a position of past Chair or Chair-elect?
Group 1: Samy Uthurusamy represented the group: The consensus was that 2 years is better. The terms could always be extended to increase continuity but there should be a policy in place. (Clayton Lewis, Rick Sward, Sung Shin, Jeremy Johnson, Paul Fishwick, David Wood, Samy Uthurusamy, Vicki Hanson, Ankur Teredesai)
Group 2: Pradip Bose represented the group: Half of our group have 3 year terms. There was a mixed discussion – one view is that it is too long. 1 year appointment is enough. In terms of significant discussions that we had, it seems that it takes about a year for SIG Chair to get to know the ropes and take on major initiatives. There was discussion on whether a 2 or 3 year term is best and the consensus was that it takes multiple years to accomplish goals. In regard to the challenge of putting together multiple candidate for each office, we received feedback this it is very difficult, especially for the Chair positions and a bit easier for the other positions. In terms of 3 year cycles which are mandatory versus 2 year cycles with automatic reappointment allowed, there was overwhelming majority from the group for 3 year cycle with mandatory re-election. (David Parkes, Rob Friedman, Pradip Bose, Walid Aref, Aidong Zhang, Kelly Wainwright, Joe Konstan, Erik Altman)
Groups 3, 4 and 5 discussed these questions and chose a representative to deliver their thoughts to the larger group:
SIG Membership Value
- What benefit of your SIG do people value most?
- Are there any benefits you would like to provide, but don't know how or find the logistics impractical, e.g. SIGxyz mugs?
- Are there any multi-SIG or cross-SIG benefits that would bring value?
- Do you belong to any other organizations which have benefits SIGs may emulate?
- How often do you visit your SIG's website?
Group 3: Represented by Cherri Pancake: We did the membership value. What benefit of your SIG do people value most? The obvious ones are conferences, Digital Library access, newsletter, SIG awards. There are others too. Social interactions, either via face-to-face or and via social media. Events for Women’s Networking. Doctorial Symposia and Industry Issuing challenges to the SIG community that they then respond to and prevent at the conference. With regard to benefits that you would like to supply but don’t know how or it is impractical, we found that Mid-Career awards are of interest. And one of the things that was suggested was that require that candidates be senior or distinguished members which might be a way of helping to promote ACM’s membership grade. Another was join awards with other awards and foundations outside of ACM. Another wish-list item is to have a Rising Star kind of award. With regard to multi-SIG and cross SIG benefits, when doing something like SIGGRAPH did and add a Mobile session, what one of the things is the idea of exchanging some presentations with SIGMOBILE with their conference. If you do that, how do you handle publications? Does it always count as the first? Does the author then choose which one comes out in the proceedings? The same things with cross-SIG tutorials – the same idea of inviting people form a different SIG to come do tutorials. With regard to other organizations’ benefits, one was where local chapters give annual top student awards. Industry pays for the awards and gives them. Another is Innovation in Teaching Award. A Regional Leaders Forum or A Regional Mentoring Forum. With regard to how often we visit your SIGs website, the majority said monthly. We also had the concept of having it come as a feed that whenever somebody updates something it comes to you as a news feed and so does that counts as going to the website? (Simon Harper, Cherri Pancake, Klara Nahrstedt, Wolfgang Banzhaf, Janis Sipior, Roy Want, X. Sharon Hu, Brent Hailpern)
Group 4: Represented by Henning Schulzreinne: We had a fairly long discussion as to the tradeoffs between making services, benefits and so on only available to members and whether that would be a good idea. Because the community is often larger than the SIG, if that is a value to the community should that be restricted to SIG members and I don’t think we came to any real conclusion. There was some interesting discussion on how we shouldn’t look at SIG membership as something equivalent to an AAA membership with towing benefits, but as something we really expect of professionals. We did talk about the usual incentives like discounted conference membership, which is a double edge sword because many companied pay for the higher conference fee but not for membership. We did see the SIG membership as essentially the larger value is that it provides the largest audience for value newsletters. Gateway membership was mentioned. With regard to the benefits which would be interesting would be chapter events and job listings. One had an affirmative action for travel grants. If you had been a long term member and were equally qualified as other travel grant applicants, the one with the longest membership would be selected for the grant over the short-term member. A possible member benefit that was discussed is that being a member is you get input on how the SIG is runs not just in terms of voting for the officers but it was suggested that a SIG might want to regularly poll its members on relevant topics. For the multi-SIG membership one interesting idea that was discussed is about clusters of SIGs to be able to offer mutual DL access for SIGs with overlapping technical interests. On the SIG website, one of the things that was suggested was a blog aggregator for SIG content. One additional member benefit was access to conference information beyond, papers like videos, conference notes (session notes) should be available to members only. One question which was raised on how to do automated membership for conference attendees which seems to be reasonably common as to whether we can standardize a process on RegOnline so it is handled in a uniform manner. (Paul Beame, James Allen, Yannis Ionnidis, Jeanna Matthews, Elisa Bertino, Henning Schulzreinne, Yolanda Gil, Patrick Madden)
Group 5: Represented by Renee McCauley: The first question about benefits of our SIGs we broke it into community versus membership. We said a lot of the same things others have said. Community is a big benefit. Conferences, awards programs and travel program. Our members, DL access, conference discount, SIG publications, the chapters, and our email distribution list. Some of the things that we might like to provide but we don’t know how, we would like to be able to record more things such as keynote addresses and make them available. We don’t also know how, or have the technology or have the funds to do such things. Social Networking around content would be useful. We would also like to offer ACM Transactions Publication as a membership benefit but we don’t know how to do that through ACM. In terms of multi-SIG and cross SIG benefits, we talked about facilitate conversations across SIGs to do things as multi-SIGs and we are not sure how to do that. In regard to possibly emulating other organizations benefits, someone brought up the issue of nominating fellows through a SIG in the manner of IEEE. Also, someone brought up the need for getting new member lists and lapsed members lists might be helpful. With regards to how often websites are visited, our group visits daily. (Elizabeth Churchill, Jeff Jortner, Jeremy Gibbons, Will Tracz, Naehyuck Chang, Renee McCauley, Donna Cappo, John C.S. Lui, Amber Settle)
Jeff Jortner: I am curious about the groups that were not in favor of 3 year terms, I am curious what is the largest sized SIG? SIGGRAPH switched to 3 year terms because of the size of the SIG the election process takes a considerable amount of time.
Henning Schulzreinne: where there is the possibility of a contentious election, like our mid-term election, you re-nominate the Chair and Vice Chair and so on, you don’t have to go through the whole nominations process which is fairly elaborate. That seemed to be a compromise and reduce a lot of work, provide the safety valve for the officer and the community and it has a reasonable term life in total. And it is then clear that after 4 years it is time to move on and get a different set of people.
From undetermined speaker: We switched from the 2 year automatic renewal so effectively 4 year term to three year terms so that in fact we have members of the Executive Committee who would want to re-up in a position because the idea is that people who have spent 4 years on it is too long for them to want to do it again. We are barely into it, I am just the second 3 year term person, except for last year.
Jeff Jortner: I am actually the second one into that also. What is very interesting is that if in chance that I got reelected then Scott Owen will essentially be an EC member for 12 years. Which is kind of an interesting thing because he was President for 2 terms, past president, nominations chair as long as I stay President, he stays nominations Chair. So that is one of the things we looked at – if you think you are going to be there long term you are really committed to be there for a long time..
Patrick Madden: SIGDA switched from 2 year with automatic extensions to three year and when I was done with my 2 years I was ready to be done.
Erik Altman closed the meeting.
The next meeting will be March 14, 2013 in Chicago, Illinois
Written by leading domain experts for software engineers, ACM Case Studies provide an in-depth look at how software teams overcome specific challenges by implementing new technologies, adopting new practices, or a combination of both. Often through first-hand accounts, these pieces explore what the challenges were, the tools and techniques that were used to combat them, and the solution that was achieved.
ACM Queue’s “Research for Practice” is your number one resource for keeping up with emerging developments in the world of theory and applying them to the challenges you face on a daily basis. In this installment, Dan Crankshaw and Joey Gonzalez provide an overview of machine learning server systems. What happens when we wish to actually deploy a machine learning model to production, and how do we serve predictions with high accuracy and high computational efficiency? Dan and Joey’s curated research selection presents cutting-edge techniques spanning database-level integration, video processing, and prediction middleware. Given the explosion of interest in machine learning and its increasing impact on seemingly every application vertical, it's possible that systems such as these will become as commonplace as relational databases are today.
Why I Belong to ACM
Hear from Bryan Cantrill, vice president of engineering at Joyent, Ben Fried chief information officer at Google, and Theo Schlossnagle, OmniTI founder on why they are members of ACM.