Abstract: By introducing the concept of information quantity and knowledge granularity to ordered information system，several properties and theorems are gained.In the process of attribute reduction，the author proves that the information quantity is monotone decreasing.Using information quantity to defined the significance of attribute.With the significance of attribute as heuristic information，a heuristic algorithm based on information quantity for reduction of attribute is presented.The time complexity of this algorithm is O（|A|³|U|²）.Finally，the experimental results show that this algorithm can find the reduction of an ordered information system.