Data mining for association rules and sequential patterns within data of inhomogeneous type

• Main Author: SCHWENKREIS FRIEDEMANN
• Format: Patent
• Language: English
• Published: 22.04.2003
• Edition: 7
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; DATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FORADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORYOR FORECASTING PURPOSES; ELECTRIC DIGITAL DATA PROCESSING; PHYSICS; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE,COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTINGPURPOSES, NOT OTHERWISE PROVIDED FOR

The invention relates to a computerized method for data mining for association rules and/or sequential patterns in a multitude of records. The invention is applicable to records that include a transaction-identification and at least one transaction-item with a corresponding item-type, wherein the multitude of records include transaction-items to be mined having different item-types. The method further includes a preprocessing step for transforming each record into one or more transaction-records in transaction-format. For each transaction-item to be data mined in a record, a transaction-record is generated and the transaction-record includes at least the transaction-identification of the record and an encoded transaction-item. The encoded transaction-item encodes the transaction-item and its corresponding item-type into one value. Finally the method includes a mining step wherein a state of the art data-mining technique is applied to the transaction-records for data mining for association rules and/or sequential patterns.