Interactive Recommendation of Composition Operators for Situational Data Integration
Guiling Grace Wang, Jun Fang, Yanbo Han · 2013
In the open and dynamic Internet computing environment, situational data integration applications appear to be trendy. The situational data integration requirements are often immediate and can't be totally defined in advance. Data service mashup is an effective approach to dealing with situational data integration problems to certain extent. It offers agility by providing the visual user interface to expose the data integration capabilities to enable non-professional users to solve transient data integration problems. In order to assist the non-professional users to select and link the composition operator, this paper aims to work toward an approach for interactive recommendation of the candidate composition operators. Based on the observation that many mashup patterns can be derived according to their functionalities, this paper proposes a composition operator recommendation algorithm combining the rule-based and statistics-based methods. We also experimentally demonstrated the accuracy of our approach.