Abstract: The availability and performance of data warehouse is gradually degrading with the altering requirements One of the biggest issues is that the set of materialized views is far from the optimal, so it is necessary for implementing the dynamic adjustment to match the demand of the users Since the current static algorithms are not suitable for this purpose on account of their larger space search and higher time consumption, this paper presents PMVS （preprocessor of materialized views selection）, an approach composed of three algorithms: QSDM （query set dynamic management）, CVLC （candidate view lattice construction） and CVF （candidate view filter） Of all these three algorithms, QSDM monitors the distribution of each query and determines by hypothesis test whether the query should be added into or discarded from the query set And based on the given query set, CVLC is in charge of producing candidate view set, which is proven to be sufficient and necessary for selecting the best set of materialized views As a heuristic algorithm, CVF then utilizes the character of data sparse in multi dimensional datasets to remove a part of candidate views that offer very limited contribution to the optimal solution The comparative experiment indicates that PMVS can be employed by the static algorithms to reduce effectively the amount of views beforehand, and the cost of static algorithms on space and time can be cut down to fit for online demand