Data envelopment analysis is very powerful quantitative approach that used for efficiency analysis and benchmarking in different and many applications through the world. Data envelopment analysis is applied as a black box. For this reason network data envelopment analysis is developed recently. So in this paper a framework with combining of network data envelopment analysis and decision tree for benchmarking of organization has been suggested. In this framework a model of decision tree is suggested for analysis of results and discovering hidden rules in the database. In this model first efficiency score is obtained by data envelopment analysis, then it is used as input of decision tree for discovering hidden rules in the database according to indexes. For validation of this suggested framework three subject of Isfahan municipality as Urban Services, Urban Planning and traort and trafic are used as decision making units. First, for evaluating of this subject, inputs and outputs are identified. Then two subjects of Urban Services and Urban Planning are evaluated. Results show that only three regions are efficient and others are inefficient. At the end of this paper decision rules for improving inefficient units is made.
