- From: Sai Surya Kiran Evani <evani@informatik.uni-freiburg.de>
- Date: Tue, 01 Jul 2003 14:27:11 +0200
- To: www-ws@w3c.org
Hi, I am currently trying to understand the different standards for describing web services and I have a few questions on DAML-S. I would appreciate it very much if somebody could throw some light on what I am missing. 1. Semantics of DAML-S : I was reading the papers "DAML-S: Semantic markup for Web Services" and "Analysis and Simulation of Web Services" which describe the semantics of web services described in DAML-S. I understand that situational calculus and petri-net based formalisms have been used to describe the semantics of composition of atomic web services is involved. My question is: if one takes the view that web service descriptions are viewed as sequences of wsdl operations(groundings in DAML-S terminology) then does not a simple formalism based on finite state automaton and temporal logic suffice to describe and discover the way a particular set of web services reacts to inputs and produce outputs? I am not able to understand how the formalisms based on situational calculus and petri-nets help in finding or match-making of web services. Or, is it the case that situational calculus etc overlap with the finite state machine formalisms. 2. Also, how would one describe a web service which wants to react to the kind of inputs sent to it in DAML-S ? For example, consider a web service description which awaits a login request or a registration request for invoking the rest of the web services. The service proceeds with the login procedure if it receives a login request or proceeds with the registration process if it receives a registration request. 3. Also, how does one describe a (closed) process in which mutiple roles are exchanging messages, basically the kind of processes that standrads like ebXML are targetting? It seems to me that the semantic web approach to web services helps services in identifying the data exchanged unambiguosly by confirming to agreed upon ontologies?? Any insights are greatly appreciated. Thanks & Regards, Kiran.
Received on Tuesday, 1 July 2003 08:39:17 UTC