AHP Method
If the problem was given to us, we would like to use AHP approach to Data quality evaluation. Analytical Hierarchy Process (AHP) is a multicriteria decision support method designed to select the best from a number of alternatives evaluated with respect to several criteria.It is taken by carrying out pairwise comparison from which priorities are developed which can be used in ranking the alternatives.
Criteria used in the AHP based approach to data quality assessment in banking are
1)Accuracy; 2)Completeness; 3)Consistency; 4) Timeliness; 5)Uniqueness; and 6)Validity. These criteria also have further subcriteria, for example under Completeness, sub criteria include Account Information, Customer information, Data Records and Interface files while a subcriteria under Accuracy is Business Indicator Range. All these variables are taken into consideration in ranking data quality. For ranking the data quality, a five step approach is adopted.
1)Structure the evaluation of hierarchy.(Goal level, criteria level, subcriteria level and alternative level)
2)Construct pairwise comparison matrices.
3)Obtain the priorities vector.
4)Test consistency
5)Weigh data quality weight and evaluate.
Dimension judgement matrix is used to see reciprocal relation between two criteria.
For the priorities vector, eigen vector is calculated for each relative judgement matrix and the eigen vector is normalized.This vector gives us the relative importance of each subindex to data quality.Consistency test value can be computed by taking the ratio of Consisteny index and Average random consistency index. The closer the consistency test value is to 0, the more consistent the matrix is. Otherwise the matrix has to be revised. Finally, the banking data quality is computed using weighted average method. The smaller the number is, the better the data quality.
Based on theis, data elements can be ranked and data elements to which data evaluation is required can be separated.
AHP based approach was
Criteria used in the AHP based approach to data quality assessment in banking are
1)Accuracy; 2)Completeness; 3)Consistency; 4) Timeliness; 5)Uniqueness; and 6)Validity. These criteria also have further subcriteria, for example under Completeness, sub criteria include Account Information, Customer information, Data Records and Interface files while a subcriteria under Accuracy is Business Indicator Range. All these variables are taken into consideration in ranking data quality. For ranking the data quality, a five step approach is adopted.
1)Structure the evaluation of hierarchy.(Goal level, criteria level, subcriteria level and alternative level)
2)Construct pairwise comparison matrices.
3)Obtain the priorities vector.
4)Test consistency
5)Weigh data quality weight and evaluate.
Dimension judgement matrix is used to see reciprocal relation between two criteria.
For the priorities vector, eigen vector is calculated for each relative judgement matrix and the eigen vector is normalized.This vector gives us the relative importance of each subindex to data quality.Consistency test value can be computed by taking the ratio of Consisteny index and Average random consistency index. The closer the consistency test value is to 0, the more consistent the matrix is. Otherwise the matrix has to be revised. Finally, the banking data quality is computed using weighted average method. The smaller the number is, the better the data quality.
Based on theis, data elements can be ranked and data elements to which data evaluation is required can be separated.
AHP based approach was