W3C home > Mailing lists > Public > public-sws-ig@w3.org > October 2004

RE: Ontology and Service Oriented Programming

From: IC Dept.- MIT-Maqbool Al Maimani <maqbool@oas.com.om>
Date: Wed, 27 Oct 2004 08:46:08 +0400
Message-ID: <FD196CAED4D46C4288E8F13F565BF91B0160C85D@EX_NT1.hq.omanair.com>
To: "Bing Li" <bing.li@asu.edu>, <public-sws-ig@w3.org>
Cc: "Li Bing" <lblabs@gmail.com>
Hi Bing Li
Congratulations for the Good Work!
I am starting my PhD this year on Semantic Web Services and require your advice on the possible topics that are still under research. I am planning to research in the area of how to recognize a requested services and automatically identify the same for the requested party.
Appreciate your response on the above
-----Original Message-----
From: public-sws-ig-request@w3.org [mailto:public-sws-ig-request@w3.org] On Behalf Of Bing Li
Sent: Wednesday, October 27, 2004 12:42 AM
To: public-sws-ig@w3.org
Cc: 'Li Bing'
Subject: Ontology and Service Oriented Programming
Dear all,
I would like to introduce my work on the area of Semantic Web Services. Your comments are welcome! By the way, since I will graduate soon, I appreciate so much if you can provide me with job opportunities. Please visit my research site, http://asusrl.eas.asu.edu/oso.
During my PhD program, I focus on distributed applications development and integration. My goals are to provide an approach that supports rapid distributed application development with implementation techniques transparency to developers and propose a solution to integration multiple distributed applications automatically. 
For the above purposes, it is discovered that ontology is a straightforward and convenient manner for developers to describe business logic. Moreover, based on ontology-oriented descriptions, it is possible to implement a converter that transforms them into programming codes, i.e., real distributed applications.
At present, service oriented architecture is an appropriate way to integrate heterogeneous systems in the Internet based environment. So I propose a fundamental developing component, service, to model a particular domain. Different from the current developing methodology, the service is described in the ontology oriented format since I believe this way is more convenient than traditional programming ones.
Besides developing using ontology-oriented descriptions, another important advantage of ontology and service oriented approach is that system integration can be based on the same specification. It saves many efforts. If standard ontology libraries exist for application domains, integration is processed automatically without human's interventions, such as invoking methods manually, understanding semantics, writing integrating scripts and so on.
In summary, ontology and service oriented programming provides a straightforward and machine comparable business logic description. Therefore, system development is much easier than current ones because of most implementation details are transparent to developers. Furthermore, system integration is carried out automatically because business logic is understandable based on matching, i.e., service discovery and selection.
Since 05/2001, I have partially implemented the above ideas. Compared with current popular research, such as DAML-S/OWL-S and BEPL4WS, my work has a lot of new contributions.
However, there still exist some issues which I should improve. First, it is clear that my current ontology oriented descriptions cannot cover all the business logic in the real business circumstance. For example, I haven't thought about the issue of timing. Second, usually, a typical distributed application is built into three-tier architecture. My final goal is to generate all the three tiers according to ontology oriented descriptions. But it is a difficult topic. Is it possible to generate a presentation tier, e.g., the navigation interface, which meets all types of customers' requirements? How to handle database management issues at the back end? Business logic in the middle tier should have different views for different customers? Third, the business logic understanding is based on matching. But the current matching approach is still simple with the too restricted conditions, a standard domain ontology library. However, in reality, it is not practical. Is it possible for my system to work on a non-standard domain ontology library? What constraints should be required for developers to describe their business logic in order to get the balance between a more natural business logic description and machine understanding?
If you are interested in that, you can visit my research Web page, http://asusrl.eas.asu.edu/oso. There you can find demo software package, demo tutorial and presentation slides. Two of my recent papers are attached with the email. They contain most of my current work. I submit them to the International Journal of Web Services Research. They are still in the review process.
I wish my work could be useful to your group projects! Your advice is highly appreciated. If I have an opportunity to work with you, it must improve my research and strengthen professional experiences to a great degree.
Looking forward to your replies!
Best regards,
Bing Li
Received on Wednesday, 27 October 2004 04:46:48 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 17:32:46 UTC