Data Quality : A Critical Issue

Data quality is a very critical issue for any organization.It plays a vital role in optimizing the data related resources.Organizations must take it seriously.This article discusses about the process of data quality.

Data quality refers to the usefulness of data at the right time at right place in an optimized manner.Utility and usefulness of data can vary from company to company.In a simple terms quality data is 'fit to use', it is trusted and suitable for its assigned objective.

According to Joseph M.Juran "Data are of high quality if they are fit for their intended used in operations, decision making and planning ".

It is very complex to ensure data quality if data are being used among various business enterprises and departments.So there is a need to measure data quality at entity and attribute level.
 

Measurement of Data Quality

Measurement of Data quality is done on following perspectives:
* Accuracy
* Reliability
* Consistency
* Relevance
* Completeness
* Timeliness

It depends on the organizations to prioritize the perspective according to their needs.Technical and business needs are required to ensure data quality.

SMART Approach
Data quality supports and helps to achieve organizational goals.There is a need of data quality project to ensure data quality.It should be SMART.You should do some SMART work for your data quality project.SMART is a mnemonic which stand for Specific, Measurable, Actionable, Realistic, and Time Driven.

Specific: data must be specific.It should be defined meaningfully at the lowest possible level.

Measurable: Data should be measured and monitored after defining data quality.

Actionable: Required Actions should be taken to improve

Realistic: Approach should be realistic.Data quality will not improve immediately.It needs realistic plan which work step wise.

Time Driven: It should be divided on time frame and written the practical milestones.

Discovery and Assessment

After setting the strategy and goals next step is discovery and assessment.
* First, data should be identified within the scope of the project.

* Define the data entities and their attributes.There are three definitions one is business definition, technical definition and quality definition.

A business definition tells about what the data is and why it is meaningful.
A technical definition provides field types, field sizes and relationships among the data.
While a quality definition contains desired and acceptable values according to the business and formatting rules.
For doing so, a tool can be used that is IBM InfoSphere Business Glossary .Business glossary guarantees the consistency of definitions across the project.

Assessment of Data

Actual data should be assessed after defining the data.
The data should be verified at different levels i.e column, table and cross-table level.It is necessary to judge its completeness, validity and conformity for its usage.
Infosphere Information Analyzer tool can be used for assessment and profiling.
* It facilitates a central business rule repository and
* It promotes reuse and consistency across different projects.
* It shares a metadata repository with InfoSphere Business glossary.

Reviewing of Data

After assessment, technical and business teams should review the outputs to understand the data completely.This complete understanding helps organizations to use data in its full extent.However it is an ideal situation.But organizations must try to approach it.Reviewing is a very important phase of data quality process.

The final Stage

Under this step, data quality process has to be created.It shifts the organizational data from current state to the expected state.

Data quality is a critical issue.Organizations and enterprises must understand their relevance and utility for functioning of business operations smoothly.Assessing and monitoring of data should be given proper preference in organizational processes.It would be helpful to avail minimum risk and maximum optimization of the data.
 

Note : Originally posted at exposeknowledge.com by same author.

Article Written by kumaresh


Post Your Comment