What Is Automated Reasoning?

An automatic inference engine is an engine that uses external information as input after the target engine sets the target and uses logical operations such as deductive induction to perform calculations on the target to generate conclusions based on the established pattern matching.

Automatic Inference Engine

Reasoning methods include precise reasoning and imprecise reasoning.

There are three types of reasoning: forward (or forward) reasoning, reverse (or backward) reasoning, and forward and reverse mixed reasoning.

Forward inference

Forward reasoning refers to deriving from the known facts and deriving towards the conclusion, until the correct conclusion is derived. This method is also called fact-driven. Its general process is: the system matches the prerequisites of the rules in the rule base according to the original information provided by the user. If the matching is successful, the conclusion part of the knowledge block is used as the intermediate result. Use this intermediate result to continue to match the rules in the knowledge base until the final conclusion is reached. Compared with other reasoning methods, forward reasoning is simple and easy to implement, but backtracking is often used in the reasoning process, which makes the reasoning speed slower, and the purpose is not strong, and can not be reversed.

Automatic reasoning

Reverse reasoning starts from the goal and goes back to the facts along the reasoning path. It starts from generality and gradually involves details, that is, it achieves the goal of solving larger problems by solving smaller sub-problems. Reverse reasoning collects more and more detailed evidence to prove a situation or hypothesis. When the data provided by the user exactly matches the evidence required by the system, the reasoning is successful and the hypothesis made is confirmed. Reverse reasoning is generally used to verify whether a particular rule holds. This type of reasoning is also called goal-driven. Compared with forward reasoning, reverse reasoning has a strong purpose.

Automatic reasoning machine forward and reverse mixed reasoning

The so-called forward and backward mixed reasoning refers to the reasoning that is based on given insufficient raw data or evidence, and then draws possible conclusions, then uses these conclusions as hypotheses, performs reverse reasoning, and finds the facts or evidence that support these assumptions . Forward and reverse mixed reasoning is generally used in the following situations:

(1) Insufficient conditions are known, and no rule can be triggered by forward reasoning;

(2) The credibility of the results obtained by forward reasoning is not high. Use reverse reasoning to solve more exact answers;

(3) Check if there are other conclusions from known conditions.

The forward and reverse mixed reasoning concentrates the advantages of forward and reverse reasoning. It is more similar to the thinking mode of people in daily decision-making. The solution process is easier to understand, but its control strategy is more complicated than the first two This method is often used to solve complex problems. ^[1]