Ontology Creation for Training and Search
Quickly grasp key ideas, players and jargon in a domain. Automatically expand search queries to retrieve relevant documents.
Domain knowledge is presented as a collection of text documents: publications, reports, textbooks, etc. In most cases, reading the entire collection is not practical or necessary. Browse through the content intelligently by building an ontology with Jaguar™.
The Lymba Solution
Jaguar™ Ontology Builder: Understand the concepts at a glance
Solution
Our automatic ontology extraction tool Jaguar™ was applied in a broad range of domains:
Law enforcement: illicit drugs, weapons, missiles, terrorism.
Finances: banking, investment, acquisitions.
Education: convert textbooks into ontologies to provide students with browsing and advanced search and to identify gaps in course materials.
Input
A collection of existing (personal) documents, Wikipedia and web pages. For a more targeted knowledge information, an existing ontology, and seeds (specific key words) can be added as an optional input.
An ontology in a standard OWL format, each concept and relation is linked to a text fragment from which it was extracted.
Output
Key Features
Automatic Generation:
Extraction of concepts and their semantic relations from a collection of documents:
– IS-A
– PART-WHOLE
– LOCATION
– RELATED
Multi-Ontology Support:
Merging and extending existing customer-provided ontologies.
Configurable:
Full control over coverage versus relevance. When ontology is required for semantic search, prefer coverage and extract maximum knowledge. On the contrary, when ontology is for personal use, it needs to be compact and contain only the most important concepts – Jaguar™ will maximize relevance.
Integration Options:
Hassle-free integration of generated ontologies into other tools like Collaborative High Precision Search (CHiPS™) our intelligent search system to support query expansion.
Interactive Browsing:
A browsing tool that allows viewing and manual curation of ontologies.