Visualising Threats

[Jo Vincent]

Australia eyes US Jigsaw system to track terrorism

Jennifer Foreshew | June 17, 2008 | Australian IT

AUSTRALIAN intelligence officers have been given an insight into new techniques, developed in the US, capable of sifting large amount of computer data to identify terrorism and organised crime threats.

Visiting US academic John Stasko, of the Georgia Institute of Technology, briefed a group of intelligence officers from the Crime and Misconduct Commission, Queensland Police Service and the Australian Federal Police in Brisbane last week.

Professor Stasko's team has developed the computer-based Jigsaw system, originally built for the US Department of Homeland Security, which allows analysts to visually map people, organisations and places to highlight connections.

Jigsaw is still a research prototype, but Professor Stasko is hoping a trial version will be used by other organisations in the next year or two.

It could be used in Australia for investigations.

Professor Stasko spoke at a seminar as part of the Griffith University Centre for Excellence in Policing and Security's inaugural crime and intelligence analysis course.

In 2007, the Australian Research Council established a Centre of Excellence in Policing and Security to support the government's Safeguarding Australia national security project.

Professor Stasko addressed the Defence Science and Technology Organisation in Adelaide during his visit.

Jigsaw can be used by investigators in law enforcement, policing, intelligence communities and reporting, particularly if there are text documents, such as case reports, news articles or blogs.

The number of documents that can be examined at one time depended on a range of factors, such as the size of the documents, but a set of about 5000 documents was considered a reasonable ceiling.

"Jigsaw doesn't do any kind of data gathering," Professor Stasko said.

"It assumes you are starting with the text documents and it will look at them to identify the entities in the documents - people, places, organisations and licence plate numbers - and to do that the tool is looking at sentences, trying to find names and so on."

Jigsaw builds a data structure of all the documents and entities and the connections between them and presents visualisations of them. There was nothing else that had the same level of visualisation in the user interface of the system, he said.

"A component of it is a search engine, so you can just search for anybody named John, or Brisbane, or whatever," Professor Stasko said.

"Once you have some documents with that, you can explore them through what is in the documents to see what is related.

"Jigsaw tries to give you a better pair of glasses, but ultimately in this visual analytics area it is still the human who is making the judgment. We want to give them a better set of tools to work with."

In the past three years, Georgia Tech, in partnership with the University of North Carolina in Charlotte, have received $US800,000 to $US900,000 in funding for a regional visualisation and analytics centre.

Professor Stasko said work was under way on a collaborative version of Jigsaw.

"Two investigators could be working on an investigation, but at totally different locations or maybe in different countries, but they are co-operating and at the same time looking at the same data."

[Jo Vincent]

ATO tunes systems to track tax cheats

Mahesh Sharma | June 03, 2008 | Australian IT

THE Australian Taxation Office will ramp up its crackdown on tax haven cheats by conducting more in-depth analysis of reports and data from the Australian financial transaction regulator.

The ATO is running a series of pilots exploring how systematic and detailed analysis of data from the Australian Transaction Reports and Analysis Centre (AUSTRAC) can improve its risk assessment of specific tax havens.

Details of the trials emerged in an Australian National Audit Office (ANAO) report released last week into the ATO’s strategies to address tax haven compliance risks.

The ATO agreed with the report’s three recommendations, including improving its assessment of the extent of compliance risks through greater trend analysis of qualitative and quantitative data.

The Department already analyses AUSTRAC data to monitor financial flows into and out of Australia, profile geographical areas, and identify potential high-risk transactions.

The report said the Department was able to match 47 per cent of AUSTRAC tax haven transaction details with Tax Office information, which was ‘very promising’ considering the data doesn’t include any ATO identifiers.

However, the ANAO said significant resources and expertise are required to boost this figure.

The ATO uses a combination of different technologies to study data including analytics modelling, Teradata data warehousing, data matching and data linking. Its chief knowledge officer Phillip Hind says it is expanding the types of data to be analysed.

“Increasingly we are turning our attention to immigration data, property and lifestyle assets, offshore credit and debit card data, and case data referred by authorities in known transfer countries.

“With these data sets, we can further match and link transactions to entities and controlling parties.”

According to the Tax Office, total direct transfers to and from tax havens increased 230 per cent to $13.7 billion over the five years to 2005/2006.