Cognitum’s year got off to a good start, with an investment from the Giza Polish Ventures Fund, and it plans to apply some of that funding to building its sales and development teams, demonstrating the approaches to and benefits of semantic knowledge engineering, and focusing on big implementations for recognizable customers. The company’s products include Fluent Editor 2 for editing and manipulating complex ontologies via controlled natural language (CNL) tools, and its NLP-fronted Ontorion Distributed Knowledge Management System for managing large ontologies in a distributed fashion (both systems are discussed in more detail in our story here). “The idea here is to open up semantic technologies more widely,” says CEO Pawel Zarzycki.
To whom? Zarzycki says the company currently has pilot projects underway in the banking sector, which see opportunities to leverage ontologies and semantic management frameworks that provide a more natural way for sharing and reusing knowledge and expressing business rules for purposes such as lead generation and market intelligence. In the telco sector, another pilot project is underway to support asset management and impact assessment efforts, and in the legal arena, the Poland-based company is working with the Polish branch of international legal company Eversheds on applying semantics to legal self-assessment issues. Having a semantic knowledge base can make it possible to automate the tasks behind assessing a legal issue, he says, and so it opens the door to outsourcing this job directly to the entity pursuing the case, with the lawyer stepping in mostly at the review stage. That saves a lot of time and money.
The company also is engaged in another pilot effort with the same law firm to develop a system that will make it possible for legal businesses to better comply with recent requirements in Europe to remove from or not insert into legal documents certain clauses that have been determined to place too many obligations on the individual. There are thousands of such clauses, he says, “and with a semantic approach we believe that it should be achievable to really make this task automatic. The core concept is you have your agreement, you put it into the system and the system automatically, with its embedded legal knowledge, verifies your agreement against this bunch of thousands of forbidden clauses” to ensure that they either aren’t in the document or can be removed if they are. Lawyers, he said, really need a solution to this problem, because searching for such clauses just on the basis of keywords easily can miss language that has to go, or alternately retrieves too many false-positive results. “It’s not a matter of simple text comparison,” he says.
In healthcare, the company is working on a research project to improve cancer treatment with the Warsaw Cancer Care Center. The Clinical Decision Supporting System (CDSS), Zarzycki says, involves building a semantic ontology for use in lung cancer care. With the help of Fluent Editor and Ontorion. this approach will assist medical practitioners in having up-to-date access to the latest knowledge around applying treatments, bringing together information from sources such as the ESMO clinical practice guidelines, and leveraging them with consideration to each particular patient’s medical history. “In the end, the medical doctor is fully guided in a semantic way with complete information about why some particular treatment should be applied,” Zarzycki says.
There’s also room for this technology in crisis management operations. The vendor is working with Warsaw University of Technology to apply a semantic knowledge management system in government administration to help operators react correctly to emergency situations, says Cognitum CTO Pawel Kapłański. “The main technology that differentiates semantic technologies from others is the possibility to express knowledge in flexible way,” he says. “We can represent the knowledge in a coherent way using semantic ontologies. And with CNL, the experts that usually have some knowledge about crisis management aren’t from software development – they are legal or other experts – and they are allowed to enter the knowledge, or at least read the knowledge, and understand and confirm if it’s true or not.”
Such projects, Zarzycki says, are all part of the journey to learn “what is the real value out of semantic technology, and how to build the solutions on top of it, and also to mix this with Big Data technologies that is more and more already,” he says. “Big Data and semantic technology is the new wave.”