This commercially-deployed product is a software tool that enables users to find meaning from text-based documents. It automatically identifies key themes, concepts and ideas from unstructured text with little or no guidance. It can be applied to emails, blogs, transcribed voice, newsgroups etc. as well as structured written text. The innovative concept map allows users to interact with the analysis-navigating the true meaning of the text-providing a perceptive tool for intelligence analysts.
The technology is based around "Automatic Concept Selection" there is no need to manually seed concepts which potentially imposes meaning. The technology will discover the important concepts from the text automatically. Conceptually simple algorithms ensure minimal imposed meaning and application across various types of text whilst "Concept Profiling" automatically discovers concepts relevant to your interest.
From the user interface perspective, the "Interactive Concept Map" allows users to understand the nature of relationships between concepts. It provides a bird's eye view, allowing the user to navigate the results and visually adjust the depth of view according to their own preference. This map can be published via a web browser facilitating simple and rapid knowledge sharing. It also features a connection to the source data allowing the user to drill-down through the concept map's results to the underlying text. A transparent thesaurus allows the user to understand how concepts are constructed.
Text can be filtered by concepts saving time in finding the data of interest. Concepts are automatically grouped into "Themes" to facilitate easier interpretation of results. Output can be colour-coded to represent more dimensions of meaning and make results easier to visualise. The product can efficiently process one, a hundred or thousands of documents in a single analysis. It has a multi-step approach to processing data providing a clear audit trail along with a full audit log. It processes a range of document formats - including HTML, word, pdf, Excel - allowing a broad scope of analysis and has the option to automatically produce a document summary.
Language independence means it can be applied to multiple languages very easily. A command-line interface allows processes to be easily scheduled or automated, and a full API supports integration with third party applications.