Project Outline and Aims

Based on studying the data manipulation requirements of users of social network data, it is clear that a sufficiently flexible data model and expressive query language for social network data has yet to be proposed. To address this, we are working on a data model (SNDM) and query language (SNQL), founded on previous research on graph databases and graph query languages.

One frequent requirement found in SN use cases is to be able to restructure a network, creating new nodes corresponding to groups or clusters of existing nodes or to values of attributes. Traditional query languages that are capable of oid or value creation tend to have the power to express all computable queries, thereby making query evaluation computationally expensive.

We have based SNQL on traditional graph query languages, with a "lightweight" extension that allows for the node creation requirements of SN manipulation. This extension is based on second-order tuple-generating dependencies, originally proposed to capture composition of schema mappings in data exchange settings.

Funding and Staffing Details

The project is being undertaken by Peter Wood, Claudio Gutierrez (U. of Chile) and Mauro San Martín (U. La Serena).

Project publications