Understanding natural language / knowlegde extraction

 

In the course of the digital transformation of business processes, it is necessary to convert instructions, contracts and contract-like documents written in natural language into a formal, computer-executable language (code) in order to implement workflow automation or smart contracts.

 

AI based on neural networks or similar statistical methods does not achieve this, because such methods can recognize well where information can be found, but cannot bring the found knowledge completely and precisely into a computer-executable form.

 

Knowledge extraction based on semantic AI does just that: based on domain-specific knowledge models, the relevant knowledge is fully extracted in texts and stored in a HOL-based executable ontology.

 

For example, customers from the aircraft industry could be helped to convert PDF-based maintenance manuals into robot executable procedures.

 

Customers from the financial industry could be helped to convert complex contracts into ontology in order to automate contract management and simulate contract scenarios in order to identify possible weak points.

​