Ontology

Ontology is a formal description of concepts in a domain and properties of each concept describing various features and attributes of the concept. It is primarily used to:

• Share common understanding of the structure of information among people or knowledge base systems.
• Reuse of domain knowledge
• New knowledge discovery

For instance, in the Healthcare space, we find that it is not easy to get the domain experts services easily and that makes this effort of building the ontology a challenging task. This problem is omnipresent across domains. Also, products/tools available are domain/task specific. They require lot of customization/training to create ontology for various domains.

What we do in Ontology

We have built a platform which will support building ontologies across domain. We can start from scratch and iteratively build the ontology using the unstructured text of the particular domain. System will automatically extract the probable terms and syntactic relationships in terms of subject-Verb-Object and then validate the extracted information against the existing domain corpus. The auto validated output is then made available to domain expert for his review. The domain expert then prune the extracted terms/relationships. This is an iterative process which will eventually develop the domain Ontology.

The claim here is that the domain expert need not put lot of effort in pruning the extraction as after every extraction and relative mapping, the system goes to an auto validation phase. However, if the baseline domain ontology is available, the process is much simpler with a few iteration itself we can expect high F-Measure from the system. Another advantage of the system is that it is UIMA compatible, so we can adapt to new features added by Open Source community very quickly.

We build a domain independent ontology platform with additional functionalities of performing inference and reasoning.
It has an Information Extraction Engine to extract domain Terms/Relationships
The platform also has a validation workbench which will help in pruning/refining the ontology.

• Application of Platform
• Social Analytics
• Semantic Search
• Text Analytics
• Sentiment Mining
• Opinion Mining
• Fact Extraction
• Event Extraction

SQL Star has proven credentials to building Knowledge Portals as Business Solution Providers. It was a natural extension that SQL Star has gained in this domain and now has capabilities to build Ontology Applications for Healthcare domain. Our team is engaged building knowledge software agents for Semantic web, Healthcare and Biomedical fields for enterprise decision support and management.