Solution

Defence / Government

The Problem

Our client authors and publishes thousands of reports each year for customers in the defence sector.

Historically, the different types of items contained within those reports remained static when a dynamically updated solution was required.

Currently, the task was labour intensive and finding the key information for customer report generation took significant time.

Our client was conscious of human error risk created through inaccurate tagging and their dependence on key personnel.

Our Solution

Adarga took samples from our client’s dictionary of terms and automatically tagged and continually updated significant volumes of forms and free form text. As our software understands the context and meaning of words and sentences to the highest possible accuracy, our client is now able to find and prioritise key data and enhance and secure corporate memory.

This involved:

  • The automation of the collection and analysis of large, unstructured data sets
  • Market specific natural language processing to accurately identify and disambiguate complex items appearing in their data
  • Entity, event and relationship extraction and corroboration
  • Summarising long documents in a meaningful way
  • Visualise relationships between individual entities and whole data sets.

The Results

80x

greater data coverage

  • 24 times faster discovery
  • 5000+ days ‘human’ reading time completed in the first 6 months
  • Up to date, automated and accurate report generation