Owing to the curse of inflated data and deficient information knowledge,we presented a new pproach to construct data warehouse architecture for supporting data analysis and decision making with object-oriented and CWM(Common Warehouse Meta model) technologies based on comprehensive analysis of current situation of metal mine enterprises in our country.Firstly,basic data of metal mine enterprises were analyzed in distribution from data statistics and production operation two points of view.The former are targets for data mining arithmetic; the latter are the members of data marts built by distinct duty of department.Then,in the light of strategies of object-oriented data warehouse,it is reasonable to go through three steps of source data area,the staging area and the presentation area under a unified spatiotemporal and dimensional framework.According to this,new dataflow architecture was put forward with CWM to implement both transaction database for metal mine daily management and data warehouse for decision making.Obeying unified modeling anguage (UML),we built concept model of metal mine entities from real world to information world,which was subsequently mapped into dimensional model of data warehouse correspondingly epending on imensional mapping rules.It shows that: for metal mine having tremendous multidimensional mineral entities,it displays more technical advantages of data warehouse of smart mine with objectoriented technique and CWM than structuring technique,so is for data integration.