公开(公告)号：GB2503622A

公开(公告)日：2014-01-01

申请号：GB201318442

申请日：2012-05-25

Applicant: IBM

Inventor： WURST MICHAEL ,  HECHLER EBERHARD ,  OBERHOFER MARTIN ,  DANTRESSANGLE PATRICK

IPC: G06F17/30 ,  G06N5/02

Abstract: The invention provides for a computer-implemented method for determining first rules (401-407), where each first rule comprises a source attribute-value pair and a destination attribute-value pair.A columnar database comprises a plurality (214; 609) of columnar data structures (109, 110; 225, 226, 227), each columnar data structure being associated with one column attribute (215-224) and comprising one or more column entries (235).First data records (213, 230-234) are stored in the columnar database. There are mask data structures (320-323), and each mask data structure has the same structure as one of the columnar data structures. The mask data structures comprise one or more second attribute- value pairs. Second data records are selected as a sub-set of the first data records by intersecting the columnar data structures and the mask data structures. One of the column attributes and one value being contained in the column data structure associated with said column attribute are selected as the destination attribute-value pair. One second rule is created for each first attribute-value pair of the second data records. For each second rule, a cooccurrence count is calculated. One or more of said second rules are selected as the first rules.