An Experimental Agricultural Data Mining System

Abstract

Agriculture is an information-intensive industry from an essential point of view. Many factors such as soil, fertilizer, temperature, precipitation, sunray, etc. are all affect harvest, so that information about them is carefully investigated by expert persons in deciding agricultural activities. We thus expect to build an intelligent agricultural information system to assist the experts and to help an improvement on agricultural technologies [7]. Towards this purpose, we firstly need to provide a system which can reveal hidden relations among agricultural factors. Although traditional statistical methods have already applied to this field, we expect recent data mining technologies to bring still more fruitful results. In particular, an expert can easily examines IF - THEN style rules extracted by the typical data mining methods [1],[6], he then may give further investigations around the rules with existing knowledge