RAVE au DISCO: recomendaÃÃo automÃtica de vÃdeo como auxÃlio no processo de disseminaÃÃo do conhecimento

AUTOR(ES)
DATA DE PUBLICAÇÃO

2009

RESUMO

In recent years, we have observed a significant growth in the production of multimedia content. It is remarkable to highlight that much of such content is spread out using the Internet infrastructure. However, some characteristics of the content access have made non-trivial the process of lookup for and retrieval of content. Difficulties to access multimedia content turn out to be even more perceptible when that content is applied for human resource training in organizations. In spite of this, the utilization of the multimedia content is increasingly common nowadays, since it contributes in many senses to support the acquisition of knowledge. The crescent presence of multimedia files in corporative environments deserves emphasis. It supports the promotion and management of corporate knowledge, while forming a base for the storage of information, such as lectures, normative manuals, video conference, audio of meetings, and so on. This sort of information storage consists of the company memory, contributing to future decision making. In the sense of making corporative decisions, it is mandatory the possibility to manipulate the content in an efficient manner throughout an structure that allows the storage and retrieval of the content in a fast and simple way. There are several research works in literature that strive to contribute to the management of organizational knowledge, such as social networks, which stimulate and allow the collaborators themselves to feed the system up. This Masterâs dissertation presents the concepts involved in the development of the RAVE environment, which is a multimedia interactive environment that minimizes the difficulties encountered when retrieving multimedia content. The objectives of the RAVE environment are not only making out the content available, rather than making automatic and customized content recommendations, using also alternative forms to retrieve multimedia content, such as a keyword search that returns the exactly time frame at which that word was pronounced. The core of this work is the content retrieval mechanism, which is a nontrivial and high cost task, in terms of communication and knowledge management requirements. The main concepts associated to the development of the RAVE environment and the results obtained in the performance evaluation are presented throughout this thesis.

ASSUNTO(S)

content-based filtering engenharia de software multimÃdia, recommendation systems filtragem hÃbrida collaborative filtering filtragem colaborativa knowledge management gestÃo do conhecimentogestÃo do conhecimento hybrid filtering ciencia da computacao multimÃdia, sistemas de recomendaÃÃo filtragem baseada em conteÃdo

ACESSO AO ARTIGO

Documentos Relacionados