Visualization of information using graphical representations of context vector based relationships and attributes

摘要: A system and method for generating context vectors use in storage retrieval of documents other information items. Context represent conceptual relationships among items by quantitative means. neural network operates on a training corpus records to develop relationship-based based word proximity co-importance using technique "windowed co-occurrence". Relationships are deterministic, so that vector set has one logical solution, although it may have plurality physical solutions. No human knowledge, thesaurus, synonym list, knowledge base, or hierarchy, is required. Summary be clustered reduce searching time, forming tree nodes. Once the determined, retrieved query interface allows user specify content terms, Boolean and/or document feedback. The present invention further facilitates visualization textual translating into visual graphical representations. Thus, can explore representations meaning, apply pattern recognition skills searches.