5. P-set structure and improved algorithm based on P-set (iMEFIM)
Although leverage strengths to exploit fast HUI Efim database changes on profit but MEFIM costly for creating projection on the set X on the transaction in question to estimate the prospects of the expansion set. With a set X are considered, the number of elements need to expand the training of forces of sub Secondary element (X).
Consider the set X and z∈Secondary (X), and the data D is the transaction containing X . Considering the projection D_ (X∪z), MEFIM forced to scan the entire D again, while it is possible to identify the collective reflections as looking secondary auxiliary element (X).
definition 19. projection X on D opposite collective
database For D and set X, P-set of the set X backward projection on D is defined as follows P-set (X) = {T.id┤ | T∈D∧X ⊆T}.
For example, X = {e} Reviews, P-set (X) = {T2, T4}.
Definition 20. The projection of the set X opposite expansion of i on D
For database D and set X, P-set-ex (X, i) the projection of the set X with the first reverse on D is defined as follows Pex-set (X, i) = {T ^ '. id┤ | CD〗 〖_X∧i T'∈ ⊆T '}.
Because P-set (X) contain transactions T.id D containing X so when performing operations related to X, you can access Quick access to transactions in P-set (X) without having done sequentially each transaction on D that accuracy does not change. The effectiveness of P-set (X) is inversely proportional to the popularity of X on D.
Example. X = {e} Reviews, P-set (X) = {T2, T4}, should take into account the usefulness of X, and we directly to T2 to T4 count instead of browsing all 5 transactions and obviously effective when use P-set ({a}) lower than the P-set ({e}) by {a} appear in more transactions {e}.
5.1. IMEFIM Algorithm
Using P-set, change algorithms at line 7 iMEFIM calculated Pex-set (X, i) parallel to the rubber (X, i) and at line 3 and 5 of the procedure Search (See Figure 8 and 9).
đang được dịch, vui lòng đợi..
