A critical mass of research in a number of AI areas made the robust and comprehensive semantic technologies possible. Unfortunately, the technology currently available is mostly of two sorts:
Both sorts require a lot of training and tuning in order to get implemented in real-world systems.
Our mission is to develop and evangelize open, skillfully engineered tools, which considerably reduce the cost for implementation and use of semantic technologies.
Formal knowledge representation (KR) is about building models of the world, of a particular domain or a problem, which allow automatic reasoning and interpretation. Such formal models are called ontologies and can be used to provide formal semantics (i.e. machine-interpretable meaning) to any sort of information: databases, catalogs, documents, web pages, etc. Having better "understanding" of the information, the machines can process it in a much more efficient manner.
Imagine, for instance, a typical database populated with the information that John is a son of Mary. It will be able to "answer" just a couple of questions: Which are the sons of Mary? and Which son is John? An ontology-based system could handle much bigger set of questions, because it will be able infer that: John is a child of Mary (the more general relation); Mary is a woman; Mary is the mother of John (the inverse relation); Mary is a relative of John (a generalization of the inverse relationship); etc. Although rather simple for a human, the above facts, would remain "invisible" to a typical database and any other information system, which model of the world is limited to data-structures of strings and numbers.
Unfortunately, building ontologies and specifying the formal semantics of the data could be an extremely slow, expensive, boring, and error-prone task. We believe that advanced linguistic methods can help for easy semantic-enrichment of documents and other data and thus to enable the wide spread of ontology-based systems.
On the other hand, ontologies are crucial for many natural language processing (NLP, knowledge discovery, text mining, etc.) tasks. They are, at the same time, the source of common sense, required to support non-trivial analysis, and the periscope, necessary to interpret, understand and make use of the results. Further, the ontologies are also playing a role in the natural language generation tasks - it is impossible to generate a reasonable, redundancy-free text without a formal model of the domain, the context, and the reader.
This is how KR and NLP, two of the most prominent AI disciplines, can live in synergy, support each other, and boost the efficiency of systems, business processes, organizations, societies, and particular human beings.