PKS has been granted a US patent for a major extension in the RippleDown rules technology which allows users to write simple rules that cover complex data domains.
The technology covered by the “method and system for generating text” patents are referred to by PKS as its “grouping” technology.
In brief, the grouping technology greatly reduces the number of rules the user would need to add by creating a data structure that allows a multitude of data items to be processed by a knowledge system as a single data item. It does this without losing the information contained in the individual data items.
This allows large volumes of data to be automatically interpreted by the rules initially added, rather than requiring humans to first filter and/or to codify that information so that it can then be processed by a knowledge system.
A key innovation of the grouping technology includes the ability to automatically analyze unstructured data such as large slabs of free form text. The PKS technology can take unstructured data and automatically extract useful information from it in any suitable reporting format.
This is useful for identifying information that may otherwise not be detected through human processing because of the time and sheer volume of data involved. Sample uses include automatic detection of ticketing fraud, monitoring of IT log files to automatically to exclude false alerts and areas of pathology reporting where a vast array of factors are taken into account.