This word extends the DEA method t,o a large number DMU evaluation process, called "DEA assessment machine", for evaluating consecutive data. We identify the intersection form of the production possibility set for a given set of DMUs, which is called "training set". The process can then easily assess a newly coming DMU for its different properties, such as technical efficiency. returns to scales and evidence of congestion, by simply checking on a, set of linear inequalities. It thus provides an efficient and effective method for dealing with large number of data and can be regarded a s a complementary approach for data mining.