Rule-based expert system for predicting regeneration results

In Finland, regeneration success has been monitored by surveys and experiments. The results are, however, difficult to be used in decision making, because they are spread in a plethora of corporate databases and research reports. The objective of this study was to gather knowledge of forest regeneration into a consistent framework. A rule-based expert system for predicting regeneration success of Scots pine was constructed. The expert system was build with KAPPA-PC applications development system that supports object-oriented approach. The object model consists of two classes (i) stands and (ii) weather. The expert system predicts regeneration success and gives recommendations concerning the choice of silvicultural system and prescriptions. The knowledge was derived from research reports. To extract the appropriate knowledge from different sources, a dependency network showing the most important factors affecting the stand density was drawn. The expert system has 101 rules, most of which are based on qualitative information. In the future, the development work will focus on quantitative rules.