Accelerating investigative discovery

Date

Authors

Flanagan, Martin

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

Since the beginning of the computing era, a key problem has been how to let 'ordinary' people ask arbitrarily deep and meaningful questions of large collections of data without being forced to resort to programming experts. The historical 'solution' was to have teams of programmers writing 'applications' for 'users' - an expensive approach whose deliverable becomes rapidly out of date leading to the well known silo problem. The Semantic Web is purposed to replace the increasingly cumbersome nature of this model - using Ontologies to describe concepts and their relationships, questions can be now phrased in terms that represent the questioners' domain of expertise in familiar language, with software then translating that into the computer code needed to retrieve the answers. The Semantic Discovery System (SDS) is InSilico Discovery's software product that implements this 'simple questions/relevant answers' Semantic Web vision, and is especially acclaimed for its unique ability to efficiently retrieve the answers from multiple distributed and disparate sources of the organisation's internal production data - as well of course as external Web sources. Gartner call this the Corporate Semantic Web - Organisations understandably want to leverage the huge value existing in production systems, but without being forced to do a mass data migration to a new architecture - they want a "Semantic Web Bridge". This bridge is the value SDS supplies - the organisational data remains in situ but it can now fulfil two purposes simultaneously - continuing to serve day to day production needs but also now supporting ad hoc research queries. SDS achieves this capability by using Semantic Web technologies (OWL, SPARQL, RDF etc) to represent a logical view of the world coupled with a Federated Query system to retrieve the physical data that will always reside in situ inside databases, files, proprietary systems etc. SDS has been built over a 10 year period (in collaboration with Universities of Pennsylvania, Manchester and GSK) based on referenceable implementations at major Pharmaceutical companies. This paper discusses capabilities of the SDS product family and plans for the immediate future.

Description

Keywords

APSR, Australian Partnership for Sustainable Repositories

Citation

Source

Type

Conference presentation

Archives Series

Date created

28/06/2007

Access Statement

Open Access

License Rights

DOI

Restricted until