Privacy-Preserving Data Mashup

Data Mashup is a special class of mashup application that combines information on the fly from multiple data sources to respond to transient business needs. Mashing up data requires an important programming skill on the side of mashups' creators, and involves handling many challenging privacy and security concerns raised by data providers. This situation prevents non-expert users from mashing up data at large. In this paper, we propose a declarative approach for mashing-up data. The approach allows data mashup creators to create data mashups without any programming involved, they just need to specify "declaratively" their data needs. The approach will then build the mashups automatically while taking into account the data's privacy and security concerns.