公开(公告)号：GB2341700A

公开(公告)日：2000-03-22

申请号：GB9911638

申请日：1999-05-20

Applicant: IBM

Inventor： EICHSTAEDT MATTHIAS ,  LU QI ,  TENG SHANG-HUA

IPC: G06F17/30

Abstract: A method for parallel matching a user profile with desired data comprises partitioning a profile database into sub-partitions having data subsets, the subsets comprising predicates used to assert selected properties to information items, mapping each sub-partition onto one or more processors yielding greatest processing efficiency, communicating an information item to each processor, and matching the information with a corresponding predicate. The mapping of the sub-partition to a processor may be based on a sub-partition's cost and the speed of the processor. If the sub-partitions have any overlap, greedy mapping partitioning may be used. Alternatively, b-mapping or cluster partitioning may be used. The method may include building a profile index (504) which includes a collection of user profiles, and which identifies predicates shared by user profiles. The method may be applied to a large scale webcasting system.

公开(公告)号：GB2341700B

公开(公告)日：2002-11-06

申请号：GB9911638

申请日：1999-05-20

Applicant: IBM

Inventor： EICHSTAEDT MATTHIAS ,  LU QI ,  TENG SHANG-HUA

IPC: G06F17/30

Abstract: A method and apparatus for efficiently matching a large collection of user profiles against a large volume of data in a webcasting system. The invention generally includes in one embodiment four steps to parallelize the profiles. First, an initial profile set is partitioned into several subsets also referred to as sub-partitions using various heuristic methods. Second, each sub-partition is mapped onto one or more independent processing units. Each processing unit is not required to have equal processing performance. However, for best performance results, subset data should be mapped in one embodiment where the subset with a highest cost is mapped to a fastest processor, and the next highest cost subset mapped to the next fastest processor. Where appropriate, the invention evaluates the relative subset processing speed of each processor and adjusts future subset mapping based upon these evaluations. For each information item I that needs to be matched with a profile predicate, a third and a fourth step are executed. The third step broadcasts I to all processing units, and a fourth step performs a sequential profile match on I.