Web Service Research Issues

Composition - How a Web services put together to form processes? Goal: Better tools for assisting human designers/developers and greater automation.
- Languages and Specifications - WSFL, XLANG, BPEL4WS, DAML-S.
- Static vs. Dynamic - In static, composition is established at design time before the Web process is run. In dynamic, some Web services are selected as need while the process is running.
- Matching - Find candidate Web services using a matching algorithms that compares "service requests" with "service advertisements". Matching based on keywords and or the types of inputs/outputs supported in UDDI. Semantic matching ...
- Mapping - Select/transform outputs from one Web service into inputs for another Web service. Facilitate code generation of service call from one Web service to the next.
- Global Optimization - Optimize overall properties of a Web process.
Enactment - How is the enactment/execution a Web process managed. Goal: Flexible and effective enactment of Web processes. What to do if something goes wrong.
- Infrastructure - Web servers, application servers, servlets, EJB, ...
- Centralized vs. Distributed - A centralized implementation is straightforward as the process can be coded as a script (e.g., in Perl or Python) or program (e.g., in Java or C#). Advantages of a distributed implementation include ...
- Exception Handling - A Web service should also indicate the exceptions it throws.
- Recovery - If a failure occurs at some point in the execution of a process instance, what can be done: alternative services, compensation services, backward recovery, forward recovery, ...
- Transactions - Support for atomicity (and isolation?). Commit protocols. Nested Transactions.
Performance Improvement - How can the performance of Web processes be improved. Goal: Produce better designs or adapt running Web processes.
- Performance Evaluation - Collecting timing data on each of the Web services.
- Monitoring - Observing the current state of the Web process (e.g., graphically).
- Simulation - Explore what-if questions about actual or potential Web processes.
- Quality of Service - Time (service time, waiting time, total time), Cost (invocation charge, leasing charge), Reliability (failure rate, exception rate).
Application to Bioinformatics - Bioinformatics is a complex, data intensive discipline in which integration/interoperability of data and computational services are problematics, thus making it well suited for the adoption of Web services.
- Wrapper Generation - Tools for creating wrappers for database or computational service which turn them into Web services.
- Data Integration - Ability to combine information from multiple data sources/databases. Combining multiple XML documents.