Supervised self organizing maps with fuzzy error correction
US7158957B2 · kind B2 · utility
Assignee
Inventors
Key dates
| Filing date | Nov 21, 2002 |
| Grant date | Jan 2, 2007 |
| Priority date | — |
| Expiry date | Dec 15, 2024 |
Classification
- Technology area (CPC Y)Emerging Cross-Sectional Technologies
- CPC primaryY10S707/99936
- WIPO fieldComputer technology
- WIPO sectorElectrical engineering
Abstract
A system and a method for an automated intelligent information mining includes receiving unstructured text from various text sources; extracting multiple key-phrases from the unstructured text; generating template and dynamic information contextual relation maps by mapping the extracted key-phrases to three-dimensional maps using a self organizing map, and a gaussian distribution technique. Further, the technique includes forming word clusters and constructing corresponding key-phrase frequency histograms for each of the generated contextual relation maps. Template and dynamic information three-dimensional structured document maps from the constructed phrase frequency histograms and the generated self-organizing maps. Desired information is extracted by mapping the generated dynamic information three-dimensional structured map onto the template three-dimensional structured map. A fuzzy prediction algorithm is used in possible error correction in extracting the desired intelligent information. A negative learning error correcting algorithm is used to correct the three-dimensional template relation maps depending on the fuzzy feedback.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.