Specifying Research Problems:

In our agent terminology or in technique history, a problem to be solved is a specific and justified manner where the agent has starts with the background in a given state and acts upon the environment until it got the altered state as per condition given has some pre-determined quality. The set of stages which are possible and manipulated via some sequence of reactions the agent takes is known as the research space or fill places. The continuously numbering of actions that the agent really performs is its searching the path, and the final stage is a solution if it has the needed property. There may be contains many solutions to a specific subject. If you just think of the problem you want your agent to perform in these terms, then you could write a problem solving agent which uses research. 

It is important to recognize the possibility of your operation in terms of the question which will need to be solved. For instance, there are some kind of situation which are single problems solved by searching them, e.g.., find a route on a map by using rough data. And one alternative manner that the tasks are like winning at chess, which have to be broken down or splits into sub-problems (searching for one of the best move at each stage to get answer). Other tasks can be produce by a simple manner without searching whatsoever e.g, multiplying two large numbers together-you wouldn't try to dream of searching through the number line you until came across the answer.