Natural language processing is a large field with many useful applications. OntoBroker provides the capability to go beyond normal content analysis by adding semantics on top.
As an example, think of a hospital with an accounting center having an eye on invoices to avoid a loss of money. The amount an insurance company pays for an operation may depend on the cause of disease. Let's say that the data set for a particular operation includes natural language information about the symptoms and also a diagnosis.
Now, for a concrete operation, $1000 can be billed, if A is the cause of disease, while for B as cause of disease, $2000 are allowed. Sadly the diagnosis might be unclear or wrong. If the symptoms indicate A, but the diagnosis is B, the hospital would bill more than it is allowed to. On the other hand, if the symptoms indicate B but the diagnosis is A, the hospital would loose money by not going for B.
This is where semantic content analytics comes in. The symptoms are taken from the natural language text. Together with an ontology modeling symptom, descriptions and diagnosis, reasoning can then be used to derive the most likely cause of disease or to check for inconsistencies.