Muhammed J. Al-Muhammed

Ontology-Based Constraint Recognition for Free-Form Service Requests

Automatic recognition and formalization of constraints from free-form service requests is a challenging problem. Its resolution would go a long way toward allowing users to make requests using free-form, natural-language-like specifications. In this paper, we address this challenge by offering an ontology-based, semantic-data-modeling approach to recognize constraints in free-form service requests. We encode domain information such as possible constraints and instances within a domain ontology in terms of object sets, relationship sets among these object sets, and operations over values in object sets and relationship sets. Our system recognizes the constraints in a service request by finding the domain ontology that best matches the request and then by using relationships and operations relevant to the request in the matched ontology to generate the service-request constraints. In experiments conducted with our prototype implementation, our system achieved an average of 96% recall and 99% precision.

Conceptual model based semantic web services

To achieve the dream of the semantic web, it must be possible for ordinary users to invoke services. Exactly how to turn this dream into reality is a challenging opportunity and an interesting research problem. It is clear that users need simple-to-invoke-and-use services. This paper shows that an approach strongly based on conceptual modeling can meet this challenge for a particular type of service—those that involve establishing an agreed-upon relationship, such as making an appointment, setting up a meeting, selling and purchasing products, or establishing employee job assignments. For these services, users can specify their requests as free-form text and then interact with the system in a simple way to complete the specification of a service request, if necessary, and invoke the service. Our system uses a conceptual-model-based information extraction ontology to (1) recognize the request and match it with an appropriate ontology, (2) discover and obtain missing information, and (3) establish agreed-upon, conceptual-model-constrained relationships with respect to the desired service. The paper lays out our vision for this type of semantic web service, gives the status of our prototype implementation, and explains how and why it works.

Resolving underconstrained and overconstrained systems of conjunctive constraints for service requests

Given a service request such as scheduling an appointment or purchasing a product, it is possible that the invocation of the service results in too many solutions that all satisfy the constraints of the request or in no solution that satisfies all the constraints. When the invocation results in too many solutions or no solution, a resolution process becomes necessary for agreeing on one of the solutions or finding some agreeable resolution. We address this problem by imposing an ordering over all solutions and over all near solutions. This ordering provides a way to select the best-m with dominated solutions or dominated near solutions eliminated. Further, we provide an expectation-based resolution process that can take the initiative and either elicit additional constraints or suggest which constraints should be relaxed. Experiments with our prototype implementation show that this resolution process correlates substantially with human behavior and thus can be effective in helping users reach an acceptable resolution for their service requests.

Bringing Web Principles to Services: Ontology-BasedWeb Services

Researchers are beginning to realize the potential of Web services that can use the Web as a place for information publication and access as opposed to the traditional Web- services paradigm that merely uses the Web as a transport medium. Traditional Web services can be difficult to discover, can have complex invocation APIs, and require strong coupling between communicating applications. In previous work, we presented ontology-based techniques in which users make service requests using free-form, natural- language-like specifications. This paper shows how we can use these ontological techniques to automatically create ontology-based Web services that (1) are easy for software agents to discover because they are created based on machine-processable formalisms (ontologies), (2) have invocation APIs requiring only simple read and write operations, and (3) require no a priori agreements regarding types and data formats between communicating applications. Experiments with our prototype implementation in several domains show that our approach can effectively create Web services with these characteristics.

Dynamic Matchmaking between Messages and Services in Multi-agent Information Systems

Agents do not work in isolation; instead they work in cooperative groups to accomplish their assigned tasks. In a multi-agent information system, we assume that each of the agents has and acquires knowledge. We further assume that it is important and useful to be able to share this knowledge and to provide useful knowledge sources to enable activities such as comparison shopping, meeting scheduling, and supply-chain management.