Data Stream Profiling Tool

The Data Stream Profiling Tool is a companion product to the TAT, and is also applicable to cyber crime. This software tool accepts ‘event streams’, such as text, keystrokes, and mouse events from multiple platforms and applications, such as word processors, in-browser text editors, instant messaging clients and so on. It matches data against known profiles in real-time and generates user alerts. An example of its use would be to detect the unauthorised use by one person of another person’s password whilst logging into a computer system. Although the correct password is used, the way in which it is typed in differs, and the tool is able to detect this difference.

The tool performs subject profiling in real-time by analysing the following characteristics:

• typing cadence
• character sequences, including use of function keys and editing characteristics
• linguistic features, at the word statistical level as well as grammatical and syntactic analysis. Ontologies can be added to enhance linguistic feature analysis and extraction.

The modular design of this tool allows new languages (or language packs) to be added readily. It also allows a language identification processing module to be added which automatically detects the language in which the person of interest is typing.

To download a copy of the Data Stream Profiling Tool brochure, please click here.