User-Driven Situational Service Mashups
Situational applications are created for a narrow group of users with a unique
set of needs to deal with situational and ad-hoc problems. For such kind of
situational applications, the requirements are in variety, can't be determined
in advance and may change over time. As a typical approach to build situational
applications, service mashups come into our attention.
Though there are some existing service mashup tools, it is still challenging
for those developers with no or little programming skills to develop service
mashups to solve the transient and ad-hoc business problems. This talk begins
with an interesting motivating scenario and discusses some of the research
issues including the data service modeling, end-user service programming. Some
research results and an on-going work on service recommendation are briefly
introduced. The initial work results have been implemented in a prototype
called ``data service space(DSS)'', which is briefly introduced in this talk.