Autometric®
Ontrack® Engenium® offers a fully automated and scalable ability to cluster documents into meaningful groups and name them with relevant, content-based labels. Cluster labels provide a clear indication of the contents, allowing simple navigation and discovery of the entire universe of available documents.
Autometric® leverages the Ontrack® Engenium® patented conceptual search technology to automatically classify documents by meaning rather than keyword, producing groupings with far greater information value than keyword-based clustering solutions. Topics and the documents associated with them are listed in descending order of relevance, with clear labels that are automatically generated by the clustering engine utilizing the ability of the Ontrack® Engenium® technology to "understand" the meaning behind the text. No manual training or label taxonomies are needed, reducing administrative overhead.
Integrated with Conceptual Search
Once the document set has sorted a particular document set into major themes and topics, users can search any document in any cluster through Semetric® concept search engine, for fast information retrieval that leverages the Ontrack® Engenium® ConceptSpace knowledge repository. The seamless integration of the conceptual search and automatic clustering delivers maximum results in minimum time.
Add-On to Other Search Engines
Alternatively, Autometric® automatic clustering can be deployed on a standalone basis with any third-party concept or keyword-based search engine for use as a culling and winnowing tool. In this case, both the list of thematic clusters generated by Ontrack® Engenium® and the documents identified as the most relevant can be loaded into the organization's search engine to guide user research. While this strategy lacks the benefit of integrating search and clustering, it jumpstarts the document exploration process, saves computing resources by reducing the size of the document set, and protects the organization's legacy search engine investment.
Features
- Concept-based organization of document collections without querying
- Instant thematic overview of large document sets to guide user research
- Intelligent labeling of each concept 'cluster' - not just high-frequency keywords
- Fully automated, with no need for manual training or label taxonomies
- Flexible administrator configuration, including the ability to specify the number of clusters desired, the number of labels used to describe a cluster, etc.
- The Ontrack® Engenium® conceptual clustering enables users to search documents within clusters
- Optional deployment as an add-on to any third-party conceptual or keyword search engines for use as a culling and winnowing tool
- Usable for any language, including Japanese and other double-byte languages
- Compatible with over 370 word processor, spreadsheet, presentation, graphics and database file types including PDF, ASCII, HTML, XML, RSS, SGML
- Scalability to millions of documents
