Using Cause-and-effect Knowledge to Solve Complex Problems

People are good at collecting the dots to describe complex problems, but atrocious at doing the logic to connect the dots to see what they imply. It is shown here how a computer program can help us by doing the logic to connect the dots. Cause-and-effect statements are used to describe the knowledge of the situation from which a behavior can arise. Given an effect to be explained that could arise from that situation, the Explainer program uses abduction to find all the plausible explanations for that effect. Then for each explanation, it finds by deduction all the effects that that explanation would predict. A valid explanation should predict behaviors that are observed to be true and not predict behaviors that are observed not to be true. The user conducts tests to determine which predictions are true and eliminates explanations that predict behaviors that are not true. Globalization and the power of the computer have unleashed great powers to do substantial things with worldwide consequences that we may not understand, making our social problems more intertwined, much more complex, and potentially more momentous. Now we need greater problem solving abilities to deal with the consequences of these actions. This technique can be used to resolve many different types of complex problems, including social problems that our government apparently has not been able to resolve. When these problems remain unsolved, our politicians and media tend to direct their energies to oversimplifications that lead to frustration, myths, emotions, arguments and gridlock. 1.0 What is the Explainer? The Explainer is a method implemented as a computer program that finds explanations for what causes behaviors. Many types of problems fall into this category. Given knowledge of a situation and given a behavior that could arise from that situation, the Explainer will find all the plausible explanations for that behavior within the scope of that knowledge. Knowledge of the situation is described by a collection of cause-and-effect statements. The behavior to be explained is an effect or set of effects. The explanation is a function of assumptions that would predict that behavior. A fact is treated as a type of assumption. The Explainer first helps the users to collect the cause-and-effect statements that describe the situation. They may also use and possibly modify available knowledge developed by others. Over time there will develop a library of such knowledge bases. Once it has the knowledge, it uses logical abduction [4,7] to find all the plausible explanations that would predict the behavior to be explained. Some of these explanations may predict behaviors that can be observed not to occur, and those explanations must be ruled out. It may also predict behaviors that had not yet been observed, leaving to further research to determine whether these other predictions also occur. The Explainer uses logical deduction for each explanation to determine all the behaviors that that explanation would predict. But this does not come right out of the box. It depends on the quality of the knowledge of the situation that is developed by the collaboration of knowledgeable people. Therefore, it is necessary to have a critical learning process to develop that knowledge. So given an explanation, the Explainer will produce a scenario showing the logical steps by which it developed that explanation. This scenario needs to be looked at carefully to see what can be learned from it and how the knowledge can be improved. This should be an ongoing process. 2.0 What are some of the types of problems the Explainer can deal with? Many types of problems can be stated in terms of what causes behaviors, and thus can be addressed using the Explainer, such as:.