Integração de sistemas de informações geográficas e sistema especialista visando auxiliar a tomada de decisão locacional do setor bancário

Visualizar/ Abrir
Data
2007Autor
Carnasciali, Ana Maria dos Santos
Metadata
Mostrar registro completoResumo
Resumo: Decidir sobre a melhor localização de um ponto comercial em qualquer setor é umadas tarefas mais importantes a ser tomada. No entanto, realizar uma decisão locacionalé uma tarefa complexa, pois envolve inúmeras variáveis, não só espaciais, comotambém subjetivas, simbólicas e muitas vezes opinativas. Considerando que paradecidir sobre uma localização nas quais estão envolvidas tanto variáveis espaciaiscomo variáveis simbólicas, esta dissertação apresenta a integração de um Sistema deInformações Geográficas (SIG) e um sistema especialista (SE) para auxiliar a tomadade decisão locacional voltada ao setor bancário, neste caso o HSBC Bank Brasil S/A –Banco Múltiplo, departamento de Administração de Imóveis (DAI), abrangendo omunicípio de Curitiba (PR). No desenvolvimento do SIG foi utilizado o softwareArcGIS 9.0 desenvolvido pela empresa ESRI, disponível na Universidade Federal doParaná (UFPR), por permitir a representação das variáveis espaciais, bem comofacilitar os processos de análises espaciais. No desenvolvimento do SE optou-se pelautilização do Shell Expert SINTA 1.1 do grupo SINTA (Sistemas InteligentesAplicados) do Laboratório de Inteligência Artificial (LIA) da Universidade Federal doCeará, por utilizar um modelo de representação do conhecimento baseado em fatos eregras. No desenvolvimento da integração entre os sistemas utilizou-se a linguagem deprogramação Visual Basic 6.0, por permitir a integração entre os sistemas, bem comoo desenvolvimento da interface do usuário. No desenvolvimento dos menus de ajudautilizou-se o software HelpScribble 7.6.1. Foi efetuada uma comparação entre ametodologia usada pelos especialistas do DAI com o sistema proposto a fim de avaliaros resultados, concluindo-se que a integração entre os sistemas contribui noaprimoramento da decisão locacional do setor bancário, ampliando as possibilidadesde análises espaciais e auxiliando os especialistas nas tarefas que envolvem pareceressubjetivos e muitas vezes opinativos. Abstract: Making a decision on where would be the best location to sell a product is the mostcrucial task in any market sector. However, this is a complex undertaking since itinvolves numerous variables, not only relating to space, but also relating tosubjectivity and symbolism and often relating to personal opinions. Taking intoconsideration that the decision on the location involves many variables as to space aswell as symbolic variables, this dissertation presents the integration of a GeographicalInformation System (GIS) with an Expert System (ES) in order to help with thedecision taking process of where to start a new ratailer in the banking sector, in thiscase within the HSBC Bank Brazil S/A, Multiple Banking, Department of Real EstateAdministration for the area of the city of Curitiba, Paraná. For the development of theGIS have used the software ArcGIS 9.0 of the company ESRI which is available at theFederal University of Paraná (UFPR), in order to enable a representation of thevariables relating to space as well as to facilitate the spatial analyzing processes.Within the development of the ES we choose to use Shell Expert SINTA 1.1 of thegroup SINTA (Intelligent Systems of Applications) of the Artificial Intelligence Lab(LIA) of the Federal University of Ceará, in order to use a model of representation ofknowledge based on facts and rules. As to the development of the integration of thesystems have used the programming language Visual Basic 6.0 in order to enable theintegration between the systems as well as to develop a user interface. For thedevelopment of the help menus we used the software HelpScribble 7.6.1. Acomparison was made between the methodology used of the specialists of theDepartment of Real Estate Administration with the proposed system in order tovalidate the results. The conclusion was that the integration of the systems contributedto an improvement within the decision maping process as to the location in thebanking sector, enlarging the possibilities of the spatial analysis and helping theexperts with the tasks that involve subjective estimations and often opinions.
Collections
- Teses & Dissertações [10497]