KNOWCODEC: A MODIFIED KNOWLEDGE NETWORK SYSTEM DEALING WITH LOGICAL STORAGE AND CONNECTIVITY OF INFORMATION BASE TO FROM KNOWLEDGE USING AUTONOMOUS NODES AND MULTI-LATERAL LINKS

摘要

Researchers in the field of Artificial Intelligence and Cognition have tried many models earlier, to simulate conceptual learning property, which can not only encode and retrieve knowledge efficiently but can also be consistent and scalable at all times. Till now knowledge Bases and Semantic web had been developed which mainly uses Ontologies to represent knowledge. Ontology development faces many challenges. To mention few of them here namely they cannot model events that are relationships between concepts and even they fail to distinguish between different relationships. Also, Ontologies cannot model events that change with time and the facts that change over time. Most important of all, the difference in representation of domain ontologies makes it hard to integrate domain ontologies. Semantic Web also imposes many challenges. To overcome the drawback of the Ontology development and Semantic Web, our invention put forward "KnowCoec", which is a modified Knowledge Network System dealing with logical storage and connectivity of information base to form knowledge, using autonomous nodes and multi-lateral links. Everything that exists as some information in human brain can be modelled as a node, a knowledge unit which is connected to other similar nodes using the multi-lateral links which posses the synaptic properties. These synaptic properties help in providing the descriptive, additive/subtractive, integrative, associative, inclusive/exclusive and temporal properties. The temporal properties help in modelling events that change with time and keep them as information for the system and even the facts that change with time. The system does not deal with language directly but represent the idea or the thought inside the human brain. Also as everything is represented using links so our invention has uniform representation for all domains and sub-domain.