Knowledge discovery agent system and method

• US 20060112029A1
• Filed: 05/21/2003
• Published: 05/25/2006
• Est. Priority Date: 05/22/2002
• Status: Active Grant

• Current Assignee: Digital Reasoning Systems, Inc (Smarsh Incorporated)
• Original Assignee: Digital Reasoning Systems, Inc (Smarsh Incorporated)
• Inventors: Estes, Timothy W.
• Granted Patent: US 7,249,117 B2
• Time in Patent Office: Days
• US Class Current: 706/16
• CPC Class Codes: G06N 5/022 Knowledge engineering; Knowledge acquisition
• CPC Class Codes: G16H 50/70 for mining of medical data, e.g. analysing previous cases of other patients

Abstract

A software agent system is provided that continues to learn as it utilizes natural language processors to tackle limited semantic awareness, and creates superior communication between disparate computer systems. The software provides intelligent middleware and advanced learning agents which extend the parameters for machine agent capabilities beyond simple, fixed tasks thus producing cost savings in future hardware and software platforms.

Claims

1. A software agent system capable of fully unsupervised learning of associations of natural language artifacts such as phrases, predicates, modifiers, or other syntactic forms and the learning of semantic and syntactic relationships in structured data sources such as is found in entities like relational database systems, tagged files, and XML.

2. A software agent system able to represent learned relationships in a particular form that allows the mapping between a variety of conventional data structures and languages such as arrays, vector spaces, first order predicate logic, Conceptual Graphs, SQL, typed programming languages (i.e. Java, C++).

3. That [1] is able to construct hierarchies of association across a state space of term usage that allows the interpolation of (weighted or fuzzy) mapping functions between sets of terms in particular syntactic positions.

4. That [1] and [3] lead to an emergent structure of weighted or fuzzy mapping functions between sets of terms that is a semantic structure analogous to formal semantic structures such as programming languages, modal logics, frame systems, or “

ontologies”

of objects and relationships

5. That [1] is able to learn from sensors that collect the interaction of human users with the system

6. That [1] is able to use the learning in [5] to reorganize and alter the mapping functions [1] induces from analyzed input, be it structured and/or unstructured, and align them to the natural usage of human users based on the sensors.