Previous work on view maintenance mostly focused on updating a single view at a time. However, to maintain a large warehouse whose source data changes rapidly, an overall maintenance strategy is needed. This paper concentrates on reducing the time taken for updating an entire set of related SPJ views based on the logical inference relationships among the views. Updating a view in response to the changes of the source data needs two types of computations: decision computation and refresh computation. The decision computation determines whether the source change affects the view while the refresh computation installs the source change into the view. By investigating the inference relationships among SPJ queries based on boolean expression relationships, an algorithm is developed to reduce the necessary computation for maintaining a set of views. The more related the SPJ views in the warehouse are, the better the performance of the algorithm is expected to be.