Brentwood, Tennessee – July 31, 2007 – Digital Reasoning Systems Inc., the intelligence-software innovator, today announced it has been awarded a patent in the United States for it’s unique method of machine learning and natural language processing.
The United States Patent and Trademark Office issued Patent no. 7,249,117 to Digital Reasoning Systems on Tuesday, July 24, entitled “Knowledge Discovery Agent System and Method”. The technology covered under this patent aims to provide a fundamentally new generation of machine-based natural human language applications.
With 38 claims, this far-reaching patent establishes Digital Reasoning as the premier technology trailblazer in the market to analyze unstructured data.
This breakthrough patent grants broad protection for how artificial intelligence, including neural networks, genetic algorithms, and vector space models can be used to learn the meanings of symbols – such as words, categories, or numerical values. Understanding the subtle meaning of terms in context has been one of the “Holy Grails” of artificial intelligence. Not only is Digital ReasoningÂ® fully able to accomplish this feat, it is now patented.
“In the landscape of information processing, this patent is very valuable real estate,” said Tim Estes, CEO of Digital Reasoning Systems, Inc. “Pretty much every conceivable approach to the area of ‘symbol grounding’ – the basic building block of determining what words mean – using unsupervised learning from measurements of contextual invariance of usage is covered under this patent.”
Here are just a few of the things this protected technology can do:
- Learn the meanings of words, classes of words, and other symbols based on how they are used in context in natural language
- Create and manipulate models of this “meaning” – i.e. the mathematical patterns of usage – including the detection of groups or similar categories of words or development of hierarchies or creation of relationships between words
- Improve the models based on human feedback or using other structured information after model construction
- The representation or sharing of this model or learning in an ontology, graph structure, or programming languages
“We believe this is a seminal moment in developing algorithms for understanding human language,” said Estes. “The applications,” he continued, “that will arise as a result of our patented technologies will allow machines to learn language much the way that children do and revolutionize the knowledge engineering process that is at the root of the most complicated systems. The implications for search engines, machine translation, and most knowledge-centric applications are immense. Finally, we can have broad machine understanding of what humans mean simply by reading what we say.”