Greetings all, I would like to make a proposal for authorship conventions on our software packages. As the years go by, the issue of the order in which authors are listed on software packages becomes increasingly muddled. We do not have a clear policy on how to order authors, and we do it differently from package to package.
I would like to propose the following convention, to take effect immediately and to be put in place whenever we issue a new version of any of our packages in the future. I will be listed as first author, and students will be listed in chronological order based on their first date of making a change that makes it into a released version of a package. This means that you will be considered an author if you make any change to a package that makes it into production, whether it is 1 line or 1000 lines. So, for example, NSP will in future have the following authors (I think): Pedersen, Banjerjee, Purandare, Thomson-McInnes, and Joshi (whew, I will need to double check that one - Mahesh Joshi has made a few changes that will be coming out in 0.73, which is in release now. I won't be able to make the authorship change for this version, but we'll start to have newer versions in the fall so we'll fix that there). I would propose the following - copyright notices will be changed to reflect this convention in new versions that come out, and we use that ordering of authors in documentation (README, etc.) Also, if there are demos of a system or papers that describe system internals, then all authors of the package will be listed as authors of such a presentation, regardless of whether or not they are participating in the demo or even the project at that time. Now, there is an important distinction to make here - if someone writes a paper using one of the tools, but that paper is not a system description, then it may be the case that all authors of the package are not authors on such a paper. For example, SenseClusters uses NSP quite a bit. I do not think all NSP authors should be authors on all SenseClusters papers. We can discuss this on a case by case basis. I hope that makes some sense, and seems fair. I do realize that there have been some hurt feelings in the past regarding these issues, and I have given a lot of thought to the above, and think it represents the fairest way to handle all sorts of issues. I have, in retrospect, been a bit careless with my handling of this issue, and I have had many regrets and second thoughts as a result. Please accept my apologies if you feel that you have been snubbed or not given proper credit in the past, and rest assured that it's very important to me that all students, current and former, receive appropriate credit for the work they do. So, when we have new versions of packages coming out from this point forward, I will propose a listing of authors on the mailing list, and make sure that everyone is ok with that (and everyone who should be included is included). When demo or system description papers are planned, we will discuss that with all package authors even if all of them are not involved in the writing of such a paper. BTW, the best (and perhaps only so far) example of system description paper is "The Implementation, Design, and Use of the Ngram Statistics Package" by Banerjee and Pedersen, CICLING 2003. See my publication page. Demos and their associated papers are probably fairly clear (we have had those at AAAI-04, AAAI-05, NAACL-05, ACL-06). If you have concerns, please feel free to let me know privately and I will adapt the proposal to meet those concerns. Finally, you may be wondering why I am first author. I guess there are two reasons. First, in most cases it might be true that I am the first one to at least think of the idea for a package. Granted, I often don't have much to do with implementation, so a different stanrdard is being applied to me. Second, as the years go by, I remain the point of contact and assume responsibility for the maintenance of the package, and all correspondence relating to the package, so this makes that clear perhaps. If you have another proposal, feel free to make it privately or to the group! Thanks! Ted -- Ted Pedersen http://www.d.umn.edu/~tpederse ------------------------ Yahoo! Groups Sponsor --------------------~--> <font face=arial size=-1><a href="http://us.ard.yahoo.com/SIG=12h4vabnn/M=323294.6903898.7846636.3189767/D=groups/S=1705082353:TM/Y=YAHOO/EXP=1123650953/A=2896129/R=0/SIG=11llkm9tk/*http://www.donorschoose.org/index.php?lc=yahooemail";>DonorsChoose. A simple way to provide underprivileged children resources often lacking in public schools. Fund a student project in NYC/NC today</a>!</font> --------------------------------------------------------------------~-> Yahoo! Groups Links <*> To visit your group on the web, go to: http://groups.yahoo.com/group/nlpatumd/ <*> To unsubscribe from this group, send an email to: [EMAIL PROTECTED] <*> Your use of Yahoo! Groups is subject to: http://docs.yahoo.com/info/terms/
