J. Smid (USA)
Knowledge representation, semantic mapping, incremental knowledge acquisition, context.
Networks are everywhere and network communications are ubiquitous. The human mind can be viewed as a communication network in which knowledge acquisition occurs incrementally. The precise way in which this knowledge is encoded is not well understood. Cognitive information may be coded similarly to natural language, a process which some argue is innate. Natural language interface may be the most universal interface. If we desire to build a system that acquires knowledge in a network environment, we can proceed by accepting some features of knowledge representation implied by the language. The incremental knowledge acquisition is arguably more robust and more desirable because it does not require reprogramming. One task we will discuss is the process of building a system that is capable of absorbing a novel piece of information. Further the system is able to answer questions and act using new information. In this paper, we outline modules and algorithms from different disciplines that can be useful for building a successful system. In particular, we discuss data structures and language semantic issues. We stress that the system is a part of networked environments and massive interactions are inherently involved.
Important Links:
Go Back
