Let View the MathML source be a given pattern. Then, the algorithm computes the weight of P as follows: View the MathML source. This weight is then multiplied by the support of P to obtain its weighted support. If the weighted support is not smaller than a user-specified threshold, P is regarded as a weighted frequent pattern. In particular, WFPMDS constructs its own tree structure and mines patterns within a single database scan in order to consider the characteristics of sliding window-based dynamic data streams. However, such approaches still have limitations in that they cannot be applied in the erasable pattern mining framework.