On an interactive system to support the management of industrial rubber tree (hevea) plantations

The use of remote sensing and Geographic Information System (GIS) for the needs of industrial plantations management, has occupied more and more important place during the last years. In the case of the hevea culture, this can be explained by the need expressed by agronomists to achieve development, inventory, maintenance and farming operations, on the plantations. by means of aerial images, topography. and descriptive data. Most of these operations integrate decisional procedures using models from Statistics, Operational Research and Artificial Intelligence. The DSS presented here has been designed in this context. Its main particularity rest on the fact that most of die decisional procedures implemented need a simultaneous use of the three types of data cited above. Our approach uses a modelization of the user's conceptual space in the shape of a state graph. A session in the system may be viewed as a partial exploration of that graph. The control of this exploraton is completly left to the user. Furthermore the system provides the user with a set of operators allowing him to generate new states from the current one, and a set of functions based on some decisional models for the evaluation of the generated states.