Today

Already today, we have two answers for the discussion about ETAI publication styles that was opened in yesterday's Newsletter. Also, a contribution by Sergio Brandano re ontologies of time.

Meta-Debates

Rob Miller:

Dear Erik,

First of all, I know I speak for many people when thanking you for the enormous effort you've put into the ETAI and the newsletter over the last year or so. People are clearly enjoying the newsletter a great deal, and it's really become part of the culture of the 'reasoning about actions' community.

Publishing an article in the ETAI has been a very positive experience for Tony and me. We had a great deal of useful feedback, and my initial worries about the subsequent status of our paper, i.e. whether it would be generally accepted as a respectable 'journal' paper, have proved unfounded. (I'm in a good position to judge this, because I recently changed jobs, and subsequently asked my new employers what their attitude had been towards this publication when evaluating my CV.) The question-and-answer sessions in the newsletter and in the online interactions page, aside from being useful and good fun, have provided good publicity for our work. So, in short, the reviewing and debating mechanisms you've set up have worked very well for us.

I therefore have only minor suggestions for changes and additions to the ETAI publication and reviewing process:

1. Encourage authors to include a statement in each ETAI article along the lines of:

   This article is best read in conjunction with the online interaction at http://......

2. Encourage the anonymous referees to post questions on the the interactions page during the 3 month period in which they're reviewing the article. In conventional journal reviewing, it's not unusual for referees to ask the authors (through the journal's editor) for clarification on various points, before making a final decision about the paper. I see this as a very healthy process which can easily be mirrored on the ETAI web pages.

3. At present, after the initial 3 month public interaction period, authors are invited to revise their paper if they wish before it is sent for anonymous review. But perhaps it would be better if the anonymous reviewers simply referee the original paper, in conjunction with the answers given to the questions posted online.

   Perhaps the anomymous reviewers could then have 3 options (similar to the options for many conventional journals):

   1. reject the paper,
   2. accept the paper but with 'compulsory' revision (and perhaps re-reviewing), or
   3. accept the paper with 'voluntary' revision.
   In this way, there would be a maximum of 2 versions of the paper online (the second appearing after 6 months rather than 3 months), so that version control wouldn't be too much of a problem. (We felt reluctant to post 3 versions of our paper in such a short time period, even though our 3rd referee felt that this might have given a better final product.)

As regards the two questions that you posed in the last newsletter, my guess is that different authors will have different attitudes; some will still like to make their articles as self-contained as possible, others will be happy to leave them as starting points for an online discussion. My view is that the ETAI can and should accomodate this range of preferences - hence suggestion (3) above.

Regards, Rob

Leora Morgenstern:

Erik,

I guess I've been lurking long enough, and I should finally just get out there and contribute to the Newsletter in particular on the question of novel publication styles. I may be a bit of a traditionalist with respect both to the issue of background material and to clarifications, and I'd vote for having these incorporated into the actual article. There is still a lot to be said for being able to print out an article and having a clear statement of the background problem in that article. I also believe that the way the author summarizes and presents background work is important; it sheds light on the author's perspective on the work. In the same way, I think it's important for the author to encapsulate the essentials of the dialogues in this forum, and to incorporate them into his article.

As a related point, I think the exercise of writing in a succinct way the background material and the main point of the dialogues serves to clarify the author's thoughts and is not something we want to give up.

Leora

Debates

Sergio Brandano:

In reply to Pat Hayes (ENRAC 3.5.1998)

An interval from the real-line is an ordered set of real numbers limited by its end-points, which are not necessarily included in the set.

It seems from this that the set of intervals is supposed to include open, half-open and closed intervals; is that right? (Or do you mean to say that there may be some doubt about whether a particular interval does or does not include its endpoints? If the latter, this is not the usual notion of 'interval' as used in real analysis, and you need to explain further.)

The former case is the one I meant.

You posed a good question, which may call into the present debate the possible relations between epistemological and ontological assumptions, at least within the "Features and Fluents" framework.

If we assume the epistemological assumption  K  (accurate and complete information about actions), then occurrences of actions are also supposed to give no doubtful information whether the scheduled time interval where they are supposed to be performed does or does not include its endpoints, so that the latter case from the quoted text must not hold. Probably the case may hold within "Mo".

Suppose  S  consists of intervals from the real line. Assume  <s1,t1> in A  and  <s2,t2> in B , intervals in  S . We say that  <s1,t1> < <s2,t2>  iff  t1 < s2 . The strict order relation

<   is an abbreviation for   < logical-and =/  . It follows then that for intervals,   <   implies   <   except for pointlike intervals (single-point closed intervals) since if  t1 < s2 , the intervals  <s1,t1>  and  <s2,t2>  cannot be equal unless  s1 = t1 = s2 = t2 .

  <   does not necessarily imply   <  , as in the case of  <2,5> < <5,9> , which is a valid case with respect to   <  .

You are right concerning the case whether  <s1,t1>  may be equal to  <s2,t2> , but this does not really affects the axiom of completeness and, into the slightest question, it may be easily fixed.

Suppose now that  <s1,t1> < <s2,t2> . The axiom of completeness states the existence of  xi in S  such that  <s1,t1> < xi < <s2,t2> .

Consider the closed intervals   [p, q]   and   [q, r]   with  p < q < r . These satisfy   <   and hence satisfy   <  , but there is no interval between them. Hence, your axiom is false for intervals on the real line.

The closed intervals   [p, q]   and   [q, r]  , with  p < q < r , do not fulfill the relation  <p,q> < <q,r> , hence they do not make a valid counterexample.

Best regards
Sergio