Trap's oil and gas resources assessment expert system is an applying branch of expert system technology in geology field.The first important thing for this system to be successful or not is to select an appropriate knowlege domain.It is the distinguishment and assessment of oil and gas pools that is the knowledge domain on which the system can be used on. The knowledge obtained should be made up of the model of oil and gas pools and the understanding of the pools that experts have in their minds.In order to make the assessment way and assessmet results to be accepted easily by users

we should make the thinking mode of the system like that of human geological specialists.So geological anology inference is used to simulate the thinking mode of geological specialists.Like human expert

the system should have considerable depth and extent of knowlodge.The depth is come from appropriate knowledge domain and from the expert's special and experiential knowledge.The extent is come from the wide scope and vast amounts of knowledge.In order to make the expert system have both great amounts of knowledge and nimble operation

it is proposed to adopt "Assembling and producing" mode for the establishment of the system's knowledge base.Since mest of the questions in petroleum geology are uncertain

the uncertainty reasoning would be the running mode of the system's inference engine.The principle of selecting an oil and gas accumulation model is that the model can stand for a kind of mechanism of oil and gas accumulation.Through the model

one can know well the accumulation process

the producing condition and the gsometry characteristics of oil and gas pools formed under this mechanism.