Overview

Fluxion is a full data integration framework, with associated tools to give Semantic Web citizens access to complex integrative solutions for any kind of federated data resource.

Storing information is easy. Data integration is hard. Fluxion aims to bring these two premises together in a complete package that seamlessly allows disparate knowledge bases to be exposed, integrated and queried together in situ.

Data is exposed in Fluxion by means of data publishers that represent underlying knowledge base structure as syntactic concepts in a distinct 'local' ontology for that knowledge base. Data and metadata are captured to create an access layer in the Web Ontology Language (OWL).

Once data is exposed, integration can be achieved by mapping between concepts in each data provider. Ontologies clearly help here by defining an agreed standard or terminology for representing data and metadata. Most data providers have focused on annotating data with ontology terms which makes querying and browsing operations much easier.

A certain degree of automatic mapping between ontologies is possible, but ultimately the researcher must understand the schema and interface of every data provider to extract useful information and they must know how the relevant information relates/overlaps. This will always be an issue for integration strategies. To this end, an overarching domain ontology, or set of ontologies, that represents common concepts in all exposed data providers is useful for Fluxion integration as database schemas do not necessarily represent the underlying data in an easily discernible manner. However, it is possible to map directly between to overlapping data provider schemas. In either case, the Fluxion project incorporates a new langauge to define ontology mapping rules and an engine to process these rules called Runcible. It is these rules that mark up the data schemas with the tacit information necessary to do live semantic data integration.

This is no easy task, so two user interfaces (UI) have been developed for Fluxion, one as a plugin for the popular Protege ontology editing tool and the other as a web-based Adobe Flash application for the Fluxion Portal; these UIs aim to simplify the rules construction process by providing a graphical interface.

A final layer is required to allow queries to be fed down to the data providers and results to be shown in a useful way. The Fluxion integrator, Pussycat, fulfils this role as a web-based graphical application to construct and harvest queries and responses, respectively.

Portal

The easiest way to start using Fluxion is to access our Portal which is a web frontend to all of the modules used in the Fluxion architecture. It provides a simpler stepwise inroad to data integration procedures that allow users to submit their data sources that are then exposed and integrated as semantically aware knowledge bases.

Further documentation

To read more in depth information about the Fluxion framework, please head over to the Documentation area.