Information Engineering for Socio-cognitively motivated networks

The Information Engineering for Socio-cognitively Motivated Network and Anomaly Discovery from Large Datasets (IE-SMNADLD) project is developing automatic methods to assist information analysts in discovering social networks among collections of text documents.

In order to provide timely, actionable advice at strategic, operational and tactical levels, information analysts typically search for relevant information and base their advice on judgements made about information contained in fragmented collections of textual documents. However, there exist many challenges that confront analysts in meeting their objectives. Firstly, there are too many sources of information and the sheer volume of information makes it extremely time-consuming for analysts to read all relevant documents which are of varying quality. Secondly, information analysts are typically time constrained and need to provide a “best anytime” analysis within a specific time constraint. Therefore, they have to prioritize, read and review information of high relevance based on diverse contexts. Thirdly, most available information is represented in textual forms and fragmented in time and space, which makes it very difficult to infer underlying connections between all sorts of puzzles over time.

To deal with these challenges, this project will:

• Develop methods and tools that assist analysts in (automatically) discovering social networks among free text documents;
• Develop methods and tools that assist analysts in identifying documents of greatest (contextual) value in a collection;
• Develop methods and tools that assist analysts to discover and track the evolution of social networks over time;
• Provide information analysts with a functionality production prototype that demonstrates these achievements.

This project is supported by the Australian Government,
Department of the Prime Minister and Cabinet, and
Australian Customs Service