Brief aan SION
Postscript,
Source,
Proposal
Bijgaand vindt u ons commentaar op de beoordeling
van onze subsidie aanvraag WASP -- Web Agent Support Program.
In het commentaar hebben we ons vooral gericht
op de aspecten waarin de beoordeling niet tegemoet komt
aan onze verwachtingen,
te weten Referee \#1: A.1.2, D.g, Referee \#2 C.
Teneinde het vergelijk met de beoordeling door de referees
en het oorspronkelijke voorstel te vergemakkelijken
is ons commentaar in het Engels gesteld.
Comments on referee-reports WASP -- Web Agent Support Program
Our comments primarily address the comments made
by the referees, i.e. Referee \#1 A.1.2, D.6 and Referee \#2 C.
A SCIENTIFIC QUALITY -- originality
Referee \#1 remarks that, in the proposal, there is little indication of
exact technology and architectures:
- As concerns technology, the realization of WASP
(and in particular Pamela) is based on the programming
language DLP (extended with Internet capability),
and the facilities provided by the hush library
and its extensions (both developed by SE/VU,
see refs [17,19,20] of the proposal).
- In addition, at the level of design, we exploit
DESIRE (developed by AI/VU, see refs [1,2,16,22] of the proposal).
In effect, the integration these technologies (DESIRE and DLP)
contributes, as rightly observed by Referee \#2, to the originality
of our approach.
- From an architectural point of view our approach
is basically a client-server architecture, with agents
as user-dedicated clients.
For the communication and Internet capabilities
we exploit existing technology (TCP/IP, HTTP, CORBA).
As regards the integration of the various components, this will
be done along the lines of the handler/event mechanism employed
in the hush library and its extensions.
B RELEVANCE TO SOCIETY -- technology transfer
Referee \#1 remarks that the technology transfer should be made clear.
-
Apart from scientific publications (in conferences and journals) the aim of WASP
is to create a prototype information discovery agent, Pamela.
As is often the case with technology developed for Internet, we will make
this prototype tool available through Internet, and announce it through
both the publications, mailing lists, netnews and our informal contacts with
colleagues. Feedback from early users will guide the development of subsequent
versions of Pamela.
Contacts with industry exist for applying the technology to be developed in
commercial and competitive environments. The contact with Rabo, mentioned in the
Appendix to the WASP proposal, has already resulted in a contract
between Rabo and
the VU for the application of agent technology in banking.
Referee \#2 remarks that the social significance
of the WASP proposal is strongly
dependent on the social significance of the Web.
- It is difficult to predict whether the popularity and social significance of the
Web will continue to increase,
or whether the Web will be replaced by a different
architecture.
The one thing which is certain is that world-wide global networking
will play an ever more important role in both the economic and social aspects
of society. As is evident from the listed publications, the applicants monitor
these evolutions closely, and take active part in them. In the architecture of
the tool to be developed, the existence of information items which are uniquely
addressable is important, but the exact data representation and the protocol for
accessing the objects are of little importance. The extraction of information from
objects, and the access routines, will be embedded in separate modules, which can
be replaced should the Web be superseded by another technology. The core of Pamela
would not be affected by such a change, making the WASP research effectively
independent of whether the global networking environment of the future will be
the Web as we know it, or a different network architecture.
C (4) COMPETENCE -- project groups
Referee \#1 remarks that there is little evidence
that the existing competence will be integrated in an effective way.
- Our answer to this remark can be found under D (6) PLANNING,
where the scientific coordination is outlined in more detail.
In addition, the intended post-doc (J. Engelfriet, see below)
will have such coordination as one of his tasks.
Referee \#2 places some doubts on the theoretical competence
of the participants.
- Apart from the theoretical expertise of the VU/AI group,
as detailed below, it may be remarked that the project leader
(Eliëns) worked in the area of formal semantics
(during his thesis, supervised by prof. dr. J.W. de Bakker).
The CV of de Bra also shows an interest and expertise
in foundational research.
-
Actually, both referees \#1 and \#2 have been unable to accurately evaluate
the theoretical competence
of the research groups participating in WASP, or even of just the applicants,
because the WASP proposal does not provide this information.
The list of publications attached to the WASP proposal only mentions those
selected publications which the applicants considered most relevant to WASP.
These publications do not describe the whole research area of the applicants,
let alone of their whole research groups.
Therefore, a short description of the areas of competence of the applicants is
given below, as well as a description of the competence of their research groups.
See appendix I.
It goes without saying that where needed or useful, other members of the
applicants' research groups will assist in the WASP project.
-
Candidate for the postdoc position:
The main candidate for the postdoc position in the project is J.
Engelfriet. He is an excellent AIO researcher with both formal and
applied interests and capabilities. His Ph.D. Thesis will be finished
within one and a half year and he is highly
interested in participating in this project.
C (5) COMPETENCE -- project leader(s)
Referee \#1 remarks that the managerial
capacity of the project leader
has not been demonstrated in the proposal.
-
The proposal format does not distinguish between project management and coordination
and scientific leadership.
The selection of a single "project leader" is intended to keep the project
management and coordination centralized. The coordinating project leader will
organize and supervise the frequent meetings of the researchers involved in WASP.
He will ensure that the different researchers communicate frequently and
collaborate actively on both the fundamental aspects of WASP and in the development
of the tool Pamela.
-
Scientific guidance and leadership will come from all four applicants,
without predefined hierarchy.
In this sense they will act as a scientific management board
with meetings on a regular basis. See D (6) PLANNING.
-
With respect to the managerial capability of the
project leader, it can be remarked that he has been chairman
of the Facultaire Onderwijscommissie for two years.
In addition, he has gained experience as leader and coordinator
of the DejaVu project group at the section Software Engineering
of the Vrije Universiteit.
D(6) PLANNING -- activity and coordination
Referee \#1 comments that the tasks involved are not well-defined
and that the coordination between the groups not made clear.
In addition, the features of Pamela have not been made
sufficiently explicit.
- First of all, it must be remarked that the target
system Pamela itself is explicitly intended to stimulate and
regulate the cooperation between the participating
research groups.
- In the design and development of Pamela
the following phases can be distinguished:
- Functional Design (Year 1..2)
- Prototype Realization (Year 2)
- Evaluation (Year 2..3)
- Final Design and Realization (Year 3..4)
We will specify the functional design
in DESIRE.
Also in the first year, work will be done on
components for the realization of the prototype,
in particular for retrieval capability, Internet access,
and presentation facilities.
The evaluation of the prototype realization
will be done in cooperation with the ASZ-GAK group
and Rabofacet.
- Periodically (every three months), meetings will be held
in which the OIO researchers (and later the post-doc)
and the scientific management team participate,
to ensure the coordination of ideas and activities.
Also, in the second and third year there will be frequent
(bi-monthly) meetings between the OIOs and the post-doc.
- As concerns its functionality,
the final version of Pamela should include
- search capability,
- modifiable user preferences, and
- (multimedia) presentation facilities.
A more detailed specification of its functionality, however,
must be considered as one of the tasks of the project.
WASP
WASP -- Web Agent Support Program
BIJLAGE I: Description of Research Groups
The descriptions below are meant to augment the descriptions
of the groups given in section 12.c of the proposal.
Since the referees were unable to evaluate the theoretical
competence of the applicants, the descriptions focus primarily on
the theoretical interests of the applicants and their groups.
A. Eliens -- SE/VU
Anton Eliëns finished his study theoretical
computer science in 1986 at the University of Amsterdam.
As employee of the CWI he did research on formal aspects
of expert systems and the semantics of distributed logic
programming. His thesis DLP - A language for distributed
logic programming, supervised by prof. dr. P. Klint and prof. dr.
J.W. de Bakker, has been published as a book by Wiley (1992).
In addition to the publications mentioned in the proposal,
he also published papers on the mathematical semantics of
distributed logic programming,
some in collaboration with dr. E. de Vink.
He is the originator and leader of
of the DejaVu project, consisting of two AIOs and
one scientific programmer, that deals with the various aspects of
object oriented programming and distributed hypermedia.
He is a member of the NVTI, ACM and IEEE.
J. Treur, F. Brazier - AI/VU
J. Treur completed his study and his thesis both in mathematics and
logic at the Utrecht University. F.M.T. Brazier completed her study in
mathematics and her thesis on information retrieval and user
interfaces both at the Vrije Universiteit Amsterdam.
The fundamental research of the VU/AI group addresses the following
topics:
- Formal semantics of reasoning and acting behaviour from a temporal
perspective
- Formalisation of cooperation in multi-agent systems
- Formal foundations for verification and validation of compositional
systems
- A logical theory of design
The researchers involved in these fundamental topics are: dr. F.M.T.
Brazier, drs. P.A.T. van Eck, drs. J. Engelfriet, dr. F. van Harmelen,
dr. C.M. Jonker, prof. dr. J. Treur, dr. L.C. Verbrugge,
drs. N.J.E. Wijngaards.
Formal semantics of reasoning and acting behaviour from a temporal
perspective.
Researchers: F.M.T. Brazier, J. Engelfriet, J. Treur, N.J.E. Wijngaards
Specifications for systems performing the complex types of reasoning and
acting processes can be formally interpreted in terms of temporal model
structures. The approach has been elaborated succesfully to obtain
formal semantics for specific types of complex reasoning processes, such
as: strategic reasoning and reflection in meta-level architectures,
default reasoning, hypothetical reasoning based on dynamic generation
and rejection of assumptions, design processes, compositional systems.
- Brazier, F.M.T., Treur, J., Wijngaards, N.J.E.,Willems, M., Temporal
Semantics of Complex Reasoning Tasks. In: B. Gaines, M. Musen (eds.),
Proc. 10th Banff Knowledge Acquisition Workshop, KAW'96, University of
Calgary, 1996
- J. Engelfriet, J. Treur, "A Temporal Model Theory for Default
Logic". In: M. Clarke, R. Kruse, S. Moral (eds.), Symbolic and
Quantitative Approaches to Reasoning and Uncertainty , Proceedings
ECSQARU93, Lecture Notes in Computer Science, Vol. 747,
Springer-Verlag, 1993, pp. 91-96.
- J. Engelfriet, J. Treur, "Temporal Theories of Reasoning". In: C.
MacNish, D.Pearce, L.M. Pereira (eds.), Logics in Artifical
Intelligence, Proceedings JELIA 94, Lecture Notes in Artificial
Intelligence, Vol. 838, Springer-Verlag, 1994, pp. 279-299. Also in:
Journal of Applied Non-Classical Logics 5(2), 1995, pp. 239-261.
- J. Engelfriet, H. Herre, J. Treur, "Nonmonotonic Belief State
Frames and Reasoning Frames (extended abstract)". In: C. Froidevaux,
J. Kohlas (eds.), Symbolic and Quantitative Approaches to Reasoning
and Uncertainty, Proceedings ECSQARU'95, Lecture Notes in Artificial
Intelligence, Vol. 946, Springer-Verlag, 1995, pp. 189-196.
- J. Engelfriet, H. Herre, J. Treur, "Nonmonotonic Reasoning with
Multiple Belief Sets". In: D.M. Gabbay, H.J. Ohlbach (eds.), Practical
Reasoning, Proceedings FAPR'96, Lecture Notes in Artificial
Intelligence, Vol. 1085, Springer-Verlag, pp. 331-344, 1996.
- J. Engelfriet, J. Treur, "Specification of Nonmonotonic Reasoning".
In: D.M.Gabbay, H.J. Ohlbach (eds.), Practical Reasoning, Proceedings
FAPR'96, Lecture Notes in Artificial Intelligence, Vol. 1085,
Springer-Verlag, pp.111-125, 1996.
- J. Engelfriet, J. Treur, "Semantics for Default Logic based on
Specific Branching Time Models". To be published in: Proceedings
ECAI'96, Wiley & Sons, 1996.
- J. Engelfriet, "Minimal Temporal Epistemic Logic". To be published
in: Notre Dame Journal of Formal Logic, special issue: Combining
Logics, 1996.
- J. Engelfriet, J. Treur, "Executable Temporal Logic for
Nonmonotonic Reasoning". To be published in: Journal of Symbolic
Computation, special issue: Executable Temporal Logics, 1996.
- J. Engelfriet, V.W. Marek, J. Treur, M. Truszczynski, "Infinitary
Default Logic for Specification of Nonmonotonic Reasoning". To be
published in: Proceedings JELIA'96, Springer-Verlag, 1996.
- J. Engelfriet, "Only Persistence Makes Nonmonotonicity
Monotonous". To be published in: Proceedings JELIA'96, Springer-
Verlag, 1996.
- Gavrila, I.S. and Treur, J. (1994). A formal model for the dynamics of
compositional reasoning systems, in A.G. Cohn (ed.), Proc. 11th
European Conference on Artificial Intelligence, ECAI'94 , John Wiley &
Sons, Chichester, 1994, pp. 307-311.
- Hoek W. van der, Meyer J.-J., Treur J.,
Formal semantics of temporal epistemic reflection,
In: L. Fribourg and F. Turini (ed.), Logic Program Synthesis and
Transformation-Meta-Programming in Logic, Proc. Fourth Int. Workshop on
Meta-programming
in Logic, META'94,
Lecture Notes in Computer Science, vol. 883, Springer Verlag, 1994,
pp. 332-352.
- Hoek W. van der, Meyer J.-J.Ch., Treur J.;
Temporalizing Epistemic Default Logic;
R.B. Feenstra, R. Wieringa (eds.),
Information Systems-Correctness and Reusability.
Selected papers from the IS-CORE-95 Workshop,
World Scientific Publishers, 1995,
pp. 173-190.
- Treur J.,
Temporal Semantics of Meta-Level Architectures for Dynamic
Control of Reasoning.
In: L. Fribourg and F. Turini (ed.), Logic Program Synthesis
and Transformation-Meta-Programming in Logic, Proceedings of the Fourth
International Workshop
on Meta-Programming in Logic, META'94.
Springer Verlag, Lecture Notes in Computer Science, vol. 883, 1994,
pp. 353-376.
Formalisation of cooperation in multi-agent systems.
Researchers: F.M.T. Brazier, C.M. Jonker, J. Treur, L.C. Verbrugge
A formalisation of a cooperation model based on joint intentions
(informally introduced by N. Jennings) has been developed and applied to
project coordination. A logical framework for collective commitments has
been introduced.
- Brazier, F.M.T., Jonker, C.M., Treur, J., Formalisation of a cooperation
model based on joint intentions, In: J. Muller, M. Wooldridge, N.R.
Jennings (eds.), Intelligent Agents III, (Proc. Third Int. Workshop on
Agent Theories, Architectures and Languages , ATAL'96), Lecture Notes in
AI, Springer Verlag, 1996, to appear
- B. Dunin-Keplicz, L.C. Verbrugge, Collective commitments, Proc. of the
Second Int. Conference on Multi-Agent Systems, ICMAS'96, MIT Press,
1996, to appear
Formal foundations for verification and validation.
Researchers: F.M.T. Brazier, F. van Harmelen, C.M. Jonker, J. Treur
Formal foundations have been developed within which static properties to
be verified and validated are defined at a semantic level, independent
of syntactical aspects of any specific representation language. Based on
the temporal semantics approach discussed above, some first explorations
have been performed to extend the logical foundation to dynamic
properties.
-
Harmelen, F. van and Aben, M., Applying rule-base anomalies to KADS
inference structures; In: R. Gamble (ed.), Proceedings of the IJCAI'95
Workshop on Validation and Verification of Knowledge-Based Systems,
1995.
- Treur J., Willems M.,
A logical foundation for verification.
In: A.G. Cohn (ed.), Proc. of the 11th European Conference on
Artificial Intelligence, ECAI'94.
John Wiley \& Sons, Chichester, 1994,
pp. 745-749.
- Treur J., Willems M.;
Formal Notions for Verification of Dynamics of Knowledge-Based Systems;
M. Ayel and M.-C. Rousset (eds.), Proc. of EUROVAV-95, European
Symposium on the Validation and Verification of Knowledge-Based
Systems,
ADERIAS-LIA, Chamb\'ery, France, 1995,
pp. 189-199.
A logical theory of design .
Researchers: F.M.T. Brazier, J. Treur
As a formal foundation for design tasks a logical theory of design has
been developed. The basic notions in design processes such as (partial)
design object descriptions, requirements and their relations are
formally defined in logical terms. As in design processes usually not
only design object descriptions but also requirements evolve over time,
and often strategic reasoning takes place, logical foundations for the
dynamic aspects are crucial. The dynamics of design processes as
formally defined in the logical theory of design are based on the
temporal semantics approach discussed above.
- Brazier F.M.T., Langen P.H.G. van, Treur J.;
A logical theory of design; J.S. Gero and F. Sudweeks (eds.), Proc. IFIP
WG5.2 Workshop on Formal Design Methods for CAD, Chapman and Hall, 1996,
to appear
P. De Bra - IS/TUE
Paul De Bra completed his study in mathematics at the University of Antwerp
in 1981, with a thesis on "Conditional Dependencies". In February 1987 he
obtained his doctorate in (computing) science, with a thesis on "Horizontal
Decompositions in the Relational Database Model", of which results have been
presented at several conferences, including the International Conference on
Database Theory (ICDT) and the conference on Mathematical Foundations for
Data Base Systems (MFDBS). Also the more recent publications [9] and [10] of
the WASP proposal show the competence of the applicant in fundamental research.
The shift in focus of attention towards more practical and applied research in
hypermedia and the Web dates back only about three years.
Paul De Bra is still a member of the "Vereniging voor Fundamentele Informatica"
(in Belgium) and of the NVTI
(Nederlandse Vereniging voor Theoretische Informatica).
He is also member of the ACM.
Paul De Bra has recently been appointed to full-professor, heading the Information
Systems section at the Eindhoven University of Technology. Apart from the
applicant this research group has expertise in executable specification
formalisms for information systems (prof. dr. K.M. Van Hee and
dr. ir. W. van der Aalst), software engineering (dr. L. Somers),
different formalisms for logic and their application to protocols (dr. F. Dignum),
object-oriented databases (prof. dr. J. Paredaens, dr. A. Aerts)
and hypermedia (dr. ir. G.J. Houben).
mailto:eliens@cs.vu.nl
(
Thu Sep 26 14:04:37 MET DST 1996
)
\end{document}