Abstract
The realms of this research include the study of decision-making methods over management policy to help draw management plans, the basic study of farm management planning, and the development of software to be used for decision-making and management-planning. However, this report focuses on the results of management policy decision-making using a text-mining system. This research, based on bookkeeping and farming data, presents a few charts of summarized management analysis that help to explain current management conditions as well as improving farm management as a whole. Farmers input information on the present state of management including the problems and solutions based on a management analysis summary. In this regard, standardized natural-language sentences can be used. The free sentences are processed into simple sentences so that statistics can be derived smoothly in a text-mining system using Morphological Analysis and Japanese Dependency Analysis. By cross-totaling the simplified sentences, major factors can be identified in the following categories: the present state of management, problems and solutions. Furthermore important sentences, chosen out of the simplified sentences to be the factors of solutions, were regarded as alternative solutions. After the evaluations were categorized as revenue, technology, labor or progress, the weighting of alternative solutions was calculated in the AHP (Analytic Hierarchy Process) system. In this way management policies in the resulting management plan were ranked in priority order.
