System and method for identifying qualifying data records from underlying databases

• US 20070027845A1
• Filed: 07/28/2005
• Published: 02/01/2007
• Est. Priority Date: 07/28/2005
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Dettinger, Richard, Kolz, Daniel
• Granted Patent: US 8,285,739 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/24578 using ranking

Abstract

A method, system and article of manufacture for creating queries against databases and, more particularly, for creating queries that are suitable to identify relevant information from one or more underlying databases. One embodiment provides a method of composing a query against data in a database. The method comprises receiving a query having one or more result fields and a plurality of query conditions. Each query condition defines selection criteria for selection of data records from the database. Each selected data record has data that satisfies at least one of the plurality of query conditions. Each of the plurality of query conditions has an associated weight. The method further comprises receiving an accuracy threshold for qualifying data records. The accuracy threshold defines a requested minimum accuracy of each selected data record defining a qualifying data record forming part of a valid result set returned for the query. For each selected data record, it is determined whether the accuracy threshold is satisfied with respect to a cumulative weight of those weighted query conditions that are satisfied.

Claims

1. A computer-implemented method of composing a query against data in a database, comprising:

receiving a query having one or more result fields and a plurality of query conditions, each query condition defining selection criteria for selection of data records from the database, each selected data record having data that satisfies at least one of the plurality of query conditions, and wherein each of the plurality of query conditions has an associated weight;

receiving an accuracy threshold for qualifying data records, the accuracy threshold defining a requested minimum accuracy of each selected data record defining a qualifying data record forming part of a valid result set returned for the query; and

for each selected data record, determining whether the accuracy threshold is satisfied with respect to a cumulative weight of those weighted query conditions that are satisfied.

2. The method of claim 1, wherein the query has at least one other query condition having no associated weight.

3. The method of claim 1, further comprising:

receiving user input specifying at least one of;

(i) a portion of the associated weights for each one of the plurality of query conditions; and

(ii) the accuracy threshold.

4. The method of claim 1, wherein a selected data record is a qualifying data record included with the valid result set returned for the query only if an accuracy value attributed to the data record as a result of the satisfied weighted query conditions is greater than the accuracy threshold.

5. The method of claim 1, wherein the accuracy threshold is expressed as a percentage indicative of the minimum accuracy.

6. The method of claim 1, further comprising:

for each query condition of the plurality of query conditions;

associating a weight with the query condition on the basis of predefined database settings; and

wherein a different weight is associated with at least two of the query conditions.

7. The method of claim 6, wherein determining, for each selected data record, whether the accuracy threshold is satisfied with respect to the weighted query conditions comprises:

executing the query against the data in the database, comprising;

for each data record contained in the database;

determining all query conditions of the plurality of query conditions that are satisfied by the data record;

determining an accuracy value for the data record on the basis of the weights associated with the satisfied query conditions; and

comparing the determined accuracy value with the accuracy threshold to determine whether the accuracy threshold is exceeded; and

including, as qualifying data records, with the valid result set for the query all selected data records having a determined accuracy value that exceeds the accuracy threshold.

8. The method of claim 7, wherein determining the accuracy value of each data record comprises:

for each satisfied query condition, determining an individual score for the data record on the basis of the weight associated with the query condition; and

determining a total score for the data record on the basis of all determined individual scores for the satisfied query conditions, the total score defining the accuracy value.

9. The method of claim 8, wherein determining the individual score for the data record comprises:

for a given satisfied query condition;

generating a value indicative of the data record; and

applying the weight associated with the given satisfied query condition to the indicative value to determine the individual score for the query condition.

10. The method of claim 1, wherein receiving an accuracy threshold for qualifying data records comprises:

receiving, for a first subset of query conditions of the plurality of query conditions, a first accuracy threshold; and

receiving, for a second subset of query conditions of the plurality of query conditions, a second accuracy threshold.

11. The method of claim 10, further comprising:

for each selected data record;

determining a first accuracy value on the basis of the first subset of query conditions, the first accuracy value being indicative of the accuracy of the data record with respect to the first subset of query conditions; and

determining a second accuracy value on the basis of the second subset of query conditions, the second accuracy value being indicative of the accuracy of the data record with respect to the second subset of query conditions; and

including, as the qualifying data records, with the valid result set for the query all selected data records having at least one of the first and second accuracy values that exceeds the accuracy threshold.

12. The method of claim 1, wherein each query condition comprises a condition field and an associated value;

and each result field and each condition field of the query is a logical field of a data abstraction model abstractly describing the data in the database;

whereby the query defines an abstract query, and wherein the data abstraction model is adapted for transforming the one or more logical fields of the abstract query into a form consistent with a physical representation of the data in the database.

13. A computer-readable medium containing a program which, when executed by a processor, performs operations for executing a query against data in a database, the operations comprising:

receiving a query having one or more result fields and a plurality of query conditions, each query condition defining selection criteria for selection of data records from the database, each selected data record having data that satisfies at least one of the plurality of query conditions, and wherein each of the plurality of query conditions has an associated weight;

executing the query against the data in the database, comprising;

identifying all selected data records from the database; and

applying an accuracy threshold to each selected data record to determine whether the accuracy threshold is satisfied for the selected data record with respect to a cumulative weight of those weighted query conditions that are satisfied thereby determining qualifying data records forming a valid result set for the query, wherein the accuracy threshold defines a requested minimum accuracy of each qualifying data record; and

outputting the valid result set.

14. The computer-readable medium of claim 13, wherein the accuracy threshold is expressed as a percentage indicative of the minimum accuracy.

15. The computer-readable medium of claim 13, wherein applying the accuracy threshold to each selected data record comprises:

determining all query conditions of the plurality of query conditions that are satisfied by the data record;

determining an accuracy value for the data record on the basis of the weights associated with the satisfied query conditions;

comparing the determined accuracy value with the accuracy threshold to determine whether the accuracy threshold is exceeded; and

including, as qualifying data records, with the valid result set for the query all selected data records having a determined accuracy value that exceeds the accuracy threshold.

16. The computer-readable medium of claim 15, wherein determining the accuracy value of each data record comprises:

for each query condition that is satisfied by the data record;

retrieving a weight indicating a relative importance of the query condition; and

determining an individual score for the data record on the basis of the weight associated with the satisfied query condition; and

determining a total score for the data record on the basis of all determined individual scores for the satisfied query conditions, the total score defining the accuracy value.

17. The computer-readable medium of claim 16, wherein a different weight is associated with at least two of the query conditions.

18. The computer-readable medium of claim 16, wherein determining the individual score for the data record comprises:

for a given satisfied query condition;

generating an indicative value for the data record; and

applying the weight associated with the given satisfied query condition to the indicative value to determine the individual score for the given satisfied query condition.

19. The computer-readable medium of claim 13, wherein applying the accuracy threshold to each selected data record comprises:

applying, for a first subset of query conditions of the plurality of query conditions, a first accuracy threshold to the selected data record; and

applying, for a second subset of query conditions of the plurality of query conditions, a second accuracy threshold to the selected data record.

20. The computer-readable medium of claim 19, wherein the operations further comprise:

for each selected data record;

determining a first accuracy value on the basis of the first subset of query conditions, the first accuracy value being indicative of the accuracy of the data record with respect to the first subset of query conditions; and

determining a second accuracy value on the basis of the second subset of query conditions, the second accuracy value being indicative of the accuracy of the data record with respect to the second subset of query conditions; and

including, as the qualifying data records, with the valid result set for the query all selected data records having at least one of the first and second accuracy values that exceeds the accuracy threshold.

21. The computer-readable medium of claim 13, wherein each query condition comprises a condition field and an associated value;

and each result field and each condition field of the query is a logical field of a data abstraction model abstractly describing the data in the database;

whereby the query defines an abstract query, and wherein the data abstraction model is adapted for transforming the one or more logical fields of the abstract query into a form consistent with a physical representation of the data in the database.

22. A system, comprising:

a database having data defining a plurality of data records;

a user interface for issuing a query against the database, the query having one or more result fields and a plurality of query conditions, each query condition defining selection criteria for selection of data records from the database, each selected data record having data that satisfies at least one of the plurality of query conditions, and wherein each of the plurality of query conditions has an associated weight; and

a query manager for executing the query against the database, the query manager being configured to;

receive the query;

identify all selected data records from the plurality of data records;

apply an accuracy threshold to each selected data record to determine whether the accuracy threshold is satisfied for the selected data record with respect to a cumulative weight of those weighted query conditions that are satisfied, thereby determining qualifying data records forming a valid result set for the query, wherein the accuracy threshold defines a requested minimum accuracy of each qualifying data record;

generate the valid result set on the basis of the determined qualifying data records; and

output the valid result set.