There are several common algorithms for constructing decision trees. The ID3 (Iterative Dichotomizer 3) algorithm was developed by Ross Quinlan in 1986 to solve the classification problem. This algorithm uses entropy as information content and implements a greedy strategy; at each node of the tree, starting from the root, it finds an attribute and threshold that provide the highest information gain for extension. It can handle missing data and has the ability to handle large amounts of data. The ID3 algorithm can be used to solve classification and prediction problems. It can be effective when working with data with a small number of attributes and where the attributes are categorical. In our work we apply the ID3 algorithm to the sets of data based on Z-information. Particularly in the context of decision analysis, the Z-number idea pertains to the question of information reliability. When estimating the alternatives, Z-number valuations account for the uncertainty of the experts’ opinions.