Integrated Law Enforcement / Data to Decisions CRC

About

Intelligence agencies and police forces increasingly rely on information that is generated by other agencies or commercial partners. Government information sources range from criminal histories, immigration records, fingerprint data, gun or vehicle registrations, video surveillance camera feeds, to house ownership and tax information. However, obtaining information in a timely manner and in a form that is suitable for analysis remains challenging, as analysts may not be aware that data pertaining to their interests is available elsewhere and data is often difficult to obtain across agency and system boundaries. Moreover, access to data must be managed carefully since sensitive content may be subject to legislative constraints and warrants, and replication or import into analytic tools may not be possible.

The Integrated Law Enforcement program seeks to develop, integrate and evaluate technology that will provide police forces and analysts with uniform access to integrated information derived from diverse data sources held by different agencies to aggregate data, resolve and link entities and identities, identify and react to unusual patterns and build and maintain threat models related to events and entities.

Streams

Development and Operations

Chris Newell (Project Lead)

Data to Decisions CRC

Integrated Law Enforcement

The Integrated Law Enforcement engineering team has been recently formed to work with the research streams by integrating the research outputs into a single product. This involves working with the research teams to advise on architectures and direction, and gaining an understanding of their outputs to integrate state of the art research into functionality to enhance the capability of Australian law enforcement agencies.

Federated Data Platform

Professor Markus Stumptner

University of South Australia

Integrated Law Enforcement

This stream will deliver a method of accessing a variety of different data sources that is stored in multiple disparate locations for use by analysts. Currently, there is a lack of capability to provide this in an automated fashion, hence there are a variety of manual techniques employed. As a result, retrieving data from disparate datasets is an extensive process and issues arise when combining data that is generated by proprietary products. This stream will address these issues and provide more capability to analysts such as entity linking, resolution across different datasets and the ability to run analysis on cross agency datasets while retaining data security and data ownership.

Entity Linking

Dr Jixue Liu

University of South Australia

Integrated Law Enforcement

This stream will enable identity resolutions based on relational data and text data, building entity profiles based on entity matching, search of the profiles using keywords, identity resolution services for user queries, maintenance of the entity profiles as the databases are updated and matching recommendations.

Data Curation Foundry for Federated Data Intelligence

Scientia Professor Boualem Benatallah

University of New South Wales

Integrated Law Enforcement

This stream aims to provide unified abstractions, languages and tools to support the entire data curation pipeline – all the way from data ingestion, analysis, to insight discovery and leverage. It will also provide the data access and curation foundry, upon which the federated analyses will operate. This will facilitate the federated analytic capabilities in an analyst-friendly, collaborative, assistive and reuse-powered manner.

Narrative Visualisation

Professor Bruce Thomas

University of South Australia

Integrated Law Enforcement

This stream is concerned with the development of tools to allow a user to develop multimedia-briefing presentations. In essence, presentations provide a non-linear means of presenting a set of data points and facts to validate arguments. The briefing is not a written document, but an interactive tool to provide a complete picture of how data and facts support a set of conclusions.

Qualitative - End User Evaluation

Professor Rens Scheepers

Deakin University

Integrated Law Enforcement

Benchmarking current practises and end user experiences of the current information systems is vital to law enforcement agencies. This project will benchmark factors such as the utility of the system to obtain insights, ease of use, discoverability of information, responsiveness, etc. The new tools and concepts designed by the Integrated Law Enforcement program will be assessed against that benchmark. That assessment forms part of the evaluation of D2D CRC as a whole.