Supervised self organizing maps with fuzzy error correction

作者: Sindhu Joseph , Sreedharan Venkataraman

DOI:

关键词:

摘要: A system and a method for an automated intelligent information mining includes receiving unstructured text from various sources; extracting multiple key-phrases the text; generating template dynamic contextual relation maps by mapping extracted to three-dimensional using self organizing map, gaussian distribution technique. Further, technique forming word clusters constructing corresponding key-phrase frequency histograms each of generated maps. Template structured document constructed phrase self-organizing Desired is map onto map. fuzzy prediction algorithm used in possible error correction desired information. negative learning correcting correct depending on feedback.