Aims:
  • Establish the design parameters required by ThoughtWeb for satisfactory validation of HarvestR
  • Identify and arrange a customer to serve as the basis for a case study and pilot programme
  • Identify and test integration points for HarvestR in the ThoughtWeb platform
  • Design and execute the pilot program and case study, including refining HarvestR as necessary to meet ThoughtWeb specifications
  • Deliver completed software and case study results to ThoughtWeb

HarvestR Validation

This project represents a rigorous scientific evaluation of HarvestR, and is expected to validate the technology in order to secure reseller licence with Leximancer and OEM licence to ThoughtWeb.

HarvestR is software used to filter, prioritise and locate information from large text-based datasets.

HarvestR represents a significant improvement over competing unstructured text analytics systems such as Autonomy and Scitext in its unique capability to automatically creatre a classification scheme or codebook for a subject area based on the inherent structure and content of a training corpus.

The HarvestR software is at a marketable stage of development and has been designed within a strong theoretical statistical framework.

However, it is currently lacking a rigorous scientific validation to demonstrate that HarvestR is able to enhance the quality and speed of analysis using real-world data.

This project seeks to validate HarvestR as this validation is critical to HarvestR’s commercial success given that the software is aimed at a high-level research and intelligence market.

ThoughtWeb is an established commercial “knowledge domain” platform that integrates 18 proprietary software systems which can be configured to meet specific business/enterprise requirements.

ThoughtWeb is an ideal commercial partner for HarvestR given its needs for improved unstructured data analysis OEM systems, and particularly given its existing commercial links to the policing and security sectors that HarvestR was originally designed to serve.

Project Value: 
$29 650.00
Funding: 
UniQuest Pathfinder
Date: 
2012 to 2013
Time status: 
Current
Contact: 
Dr Gentry White (g.white@uq.edu.au)
Aims:
  • Establish the design parameters required by ThoughtWeb for satisfactory validation of HarvestR
  • Identify and arrange a customer to serve as the basis for a case study and pilot programme
  • Identify and test integration points for HarvestR in the ThoughtWeb platform
  • Design and execute the pilot program and case study, including refining HarvestR as necessary to meet ThoughtWeb specifications
  • Deliver completed software and case study results to ThoughtWeb
Program/Affiliation: 
Type: 
Quantitative
Keywords: 
Resilience
Crime control
Terrorism
Intelligence Methods
Number: 
ISSR030352