A composite data service is composed of a number of basic data services with hierarchical structure.It is one of the important approaches for cross-domain data integration.Once a primitive data service is updated

it becomes a question of how to propagate the update bottom-up to the top level via several intermediate composite data services

and to ensure the update efficiency of the data view.This paper intends to present a cache-based optimization method for the composite data services

which is based on a cost-benefit model of cache points to find the best cache point of the composite data services.In addition

when the data services update

we can find the cache point quickly and efficiently to reuse the cached data through the B-tree index

that avoid repeat of the updated calculation.Experiments show that our approach can effectively improve the update efficiency of the composite data services.