To measure the structure relationship among indexes and different indexes groups based on the networkrelationship information. Methods: Tabu search algorithm was used as the concrete algorithm to study the composition ofindex groups and the structure relationship between them. Result: The results of simulation and real data show that Tabusearch algorithm accurately solve the division of indexes and measure the structure relationship among different groups.Conclusion: Tabu search algorithm measure the structure relationship among indexes and different indexes groups, whichextend to larger network and more complex situations.