Return to search

A multi-agent collaboration model for knowledge sharing

M.Sc. / Knowledge has become a valuable resource for organisations contributing to their competitive advantage through innovation. Over the years, the capture of organisational information has grown to include structured information, such as corporate databases, as well as unstructured information, such as best practices, policies, rules and strategies. However, most of the organisational tacit knowledge gained through experience and development of expertise is never captured by the organisation and remains with employees, leaving the organisation whenever employees depart. The goal of the dissertation is to understand how knowledge travels through an enterprise, in order to gain insight into the possibilities of capturing organisational tacit knowledge, while encouraging employee collaboration. Based on this understanding, the aim is to propose a model that will allow knowledge sharing and encourage collaboration in order to present the organisation with an alternative knowledge base of daily user activities, organisational experiences and learning. The proposed collaboration model makes use of multiple agents, namely a user agent, a conversation manager agent, question analysis and query agents and a conversation classifier. The model aims to store knowledge base entries in an ontology-based structure and to leverage off the power of existing search engine technologies to perform searches. An implementation of the proposed model is discussed using Microsoft Office SharePoint Server, a popular enterprise collaboration platform, which assists implementation and user adoption.
Date31 August 2011
CreatorsSloan, Nevena
Source SetsSouth African National ETD Portal
Detected LanguageEnglish

Page generated in 0.002 seconds