Summarization or aggregation is a process of computing the measures in the data cube. Summarization plays an important role in the data analysis and decision making for data mining applications. However, the potential of inaccurate summarization that could result from using improper operators may lead to inaccurate measure values that affect data analysis and the decision making process. Efficient computation of measures and an accurate summarization process have become an important requirement in order to obtain useful results that best support the decision making process. To this purpose, a new operator for summarization based on linear goal programming is proposed. The proposed operator computes new measure values with minimum distance to its related true values in order to maintain data quality and find higher accuracy measures. Higher accuracy measures reflect more useful analysis for users and improves decision making. The ability of the proposed operator to achieve higher accuracy when performing roll-up operations and compute new measure values that best reflect its correspondent original data values is measured and compared with average, minimum, and 394 Anas Jebreen Atyeeh Husain maximum aggregation operators. The evaluation results demonstrated that the goal programming operator performs better than the others and produces more accurate measure values by achieving a lower distance to the related data values. We conclude that the proposed operator is able to improve the summarization process that best supports decision making by providing higher data quality to the OLAP users.