Decision making in the forest protection involves diagnosing the pest, making predictions of the effects of the pest on forest, knowing the possible control tactics, and cost/benefit integration. To cope with all that, a generalist forest manager needs a tool like an expert system to support decisions.
This paper presents an expert system that approaches the goals of integrated pest management. With the systm, the user can make diagnosis and predictions of 12 North European bark beetles. Written in Common LISP and Flavors, the expert system has a combined frame- and rule-based knowledge representation. Frames are used to represent the hierarchy of insect taxonomy in diagnosis. Prediction is made with qualitative reasoning with rules. The interface engine applies both forward and backward chaining. The system has a graphical user interface that supports exploring the sensitivity of advice on input.
It is concluded that expert systems and artificial intelligence have high applicability everywhere in forestry where complicated decisions have to be made. Especially, an integrated pest management system in forestry is largely equivalent to a computerized decision-making aid.
The PDF includes an abstract in Finnish.