In the past decades, the tremendous growth in computer-based information technology has led to the need for an efficient handling of information systems for the purpose of supporting decision making. The information systems are combinations of hardware and software that are used for storing and processing data; several techniques were introduced to extract useful information through the concepts of topological spaces.Attribute reduction and function selection have end up one of the essential steps for pattern recognition and machine getting to know responsibilities. The purpose of attribute reduction is to discover a minimal attribute subset that satisfies a few particular criteria, even as the minimal attribute subset is referred to as attribute reduce.The purpose of attribute reduction is to discover a minimal attribute subset that satisfies a few particular criteria, even as the minimal attribute subset is referred to as attribute reduce. In this paper a pretopological space from an information system is to be constructed and examined, as well as defining some essential operations for the pretopological space and a brief study of its properties will be presented. Moreover, we are presenting a methodology for attributes reduction by gauging the accuracy degrees of decision sets in an information system.