Data Quality Improvement Strategies

 To improve the data quality, there are many cleansing tools might be useful tools for f the activities’ automation which is involved in the cleansing of adjustment, transformation, data-parsing and standardizing. Without the validation of the data models, the continuous progress in the entities cannot be done. Entities are linked to make a strong relationship between them using the ERD. It is not meaningfully devised a unique data quality concept. There are many techniques of improving data quality and to be applied the data in a specific type, and it should carefully consider the analytics over the data. RDBMS represented the data in the tabular format, and there is some sort of relation between the tables. Dimensions are usually defined as a qualitative method, it refers to the general types of data that is associated with dimensions are separate. When people asked about the data quality, they often asked about the data accuracy (Chen, Meyer, Ganapathi, Liu, & Cirella, 2011).

Data quality management has discussed the multiple techniques for assessing the quality of data along multiple dimensions but practitioners and researchers have underscored the contextual quality assessment’s importance in recent years and highlighted its contribution to the decision-making.  In the paper there would be discussion about persistent attention to contextual aspects, latest data quality measurement methods need to be revised and alternatives should be considered that reflect the contextual evaluation in a better way (Even & Shankaranarayanan, 2005).

Literature review of Data Quality Improvement Strategies

Chen et al. (2011) examined the improving quality of data in the relational database: overcoming functional entaglements. According to the authors, common data anomalies have failed to be prevented by the methods of traditional vertical decomposition in the relational database normalization. Moreover, authors have stated that data quality may still be deteriorated even after that database can be normalized highly and the potential data anomalies are the reason of this deteriorated data quality. Authors have discussed that database needs to be further improved by practitioners after applying the methods of traditional normalization, because of the functional entanglement’s existence; authors have defined this phenomenon. Two methods have been outlined in the paper for functional entanglements identification in the normalized database that would be the first place towards the improvement in data quality.

Furthermore, several practical methods have analyzed by the researchers for common data anomalies prevention by restricting and eliminating the functional entanglements’ effects. Authors have revealed the traditional method’s shortcomings in database normalization with admiration to the common data anomalies’ prevention and offered the useful techniques to practitioners for data quality improvement. Horizontal decomposition and field-level disentanglement are the two methods examined by authors for functional entanglements elimination at the normalized database’s design level. Authors have suggested that the requirements should be carefully evaluated by the practitioners to apply the most suitable method while dealing with potential data anomalies (Chen et al., 2011).

The have conducted the data quality tools' survey. According to the authors, data is transformed by data quality tools with problems into the good quality data for some specific application domain of the organization. Commercial, as well as research data quality tools’ classification, is presented in the paper according to the three perspectives i.e. data quality problems’ taxonomy that is not addressed by the current technology used in the  RDBMS, generic functionalities listing and division of data quality tools into many of the groups. Data profiling, analysis, transformation, cleaning, duplicate elimination and data enrichment are the six categories of tools identified.

According to the research conducted by Singh & Singh, (2010), it is descriptively classified the causes and problems of data quality in data storage. Authors have addressed this issue in the perspective of organizations that organizations have become aware of the decision-oriented benefits and oriented databases of business intelligence, this is why data recording keeping is gaining eminence. Populating process of data is quite a difficult task, especially with quality data and authors, have contributed the issues of data quality over the period of time but this research has gathered the data quality problem's causes collectively at the data phases such as data sources, data profiling and data integration, ETL and Data staging, schema design and database modeling. The authors have identified the reasons for reachability problems, on-availability and data deficiencies at data storage aforementioned stages such as insufficient data profiling, content analysis and automating profiling tools ‘inappropriate selection (Singh & Singh, 2010).

Batini et al., (2007) monitored as well as assess the methodology and framework for data quality. According to the authors, data quality is emerging as the new area organizations' improvement of the effectiveness. Poor quality data are frequently experienced in the routine life of enterprises despite its bad consequences and there are very few organizations that use specific methodologies for monitoring and assessing the quality of organizational data. Authors have presented the Italian project's first results whose objective is to produce the well-known approaches' enhanced version for the Basel II operational risk evaluation along with the significant relevance to the data quality and information and its impact on the operational risk. Definition of the assessment methodology is the focus of authors in this paper along with the supporting tool for data quality. Authors have explained the different phases of steps of methodology such as data quality risk prioritization, risk identification, risk measurement and risk monitoring. The methodology developed by the authors is based on the even loss’s notion caused by the low quality of data (Batini et al., 2007).

Sadiq & Duckham examined the querying and storing data quality information in the context of spatially varying by using the integrated spatial RDBMS. According to the authors, the current SDQ (Spatial Data Quality) representation do not represent the SDQ's spatial variation adequately. For instance, if some user wishes to get familiar with the feature's positional accuracy then normally has to rely on the statements of metadata that normally refers to the entire set of data. Authors have stressed that SDQ varies spatially in reality; in some location, quality time may be higher and in other locations, it can be the lower perhaps because of the different procedures of data collection as well as methods of acquisition. SDQ need to be stored to represent the data quality that varies spatially individual features as well as feature's part in the dataset.

The authors of this research have proposed the flexible and new data model for the spatially varying quality information’s retrieval and efficient storage in the spatial database. The quality information is stored in the model proposed by the authors in several ways according to the data set’s requirements. The authors have reported on the expansion to Oracle spatial RDBMS that is being used in order to implement the spatially varying SDQ's model. The authors have conducted the investigations into the several querying mechanisms that are needed to support the SDQ model, the investigations have shown that spatially varying quality’s flexible representation is allowed by the different models including the sub-feature variation in the quality (Sadiq & Duckham).

Conclusion of Data Quality Improvement Strategies

Summing up the discussion it can be said that there are many techniques of improving data quality and to be applied the data in a specific type and it should carefully consider the analytics over the data. The traditional method’s shortcomings in database normalization with admiration to the common data anomalies’ prevention and offered the useful techniques to practitioners for data quality improvement. Commercial, as well as research data quality tools’ classification, is presented in the paper according to the three perspectives i.e.  Populating process of data is quite a difficult task, especially with quality data and authors, have contributed the issues of data quality over the period of time

Poor quality data are frequently experienced in the routine life of enterprises despite its bad consequences and there are very few organizations that use specific methodologies for monitoring and assessing the quality. SDQ need to be stored to represent the data quality that varies spatially individual features as well as feature's part in the dataset. The quality information is stored in the model proposed by the authors in several ways according to the data set’s requirements.

References of Data Quality Improvement Strategies

Barateiro, J., & Galhardas, H. (2005). A survey of data quality tools. Datenbank-Spektrum, 15-21.

Batini, C., Barone, D., Mastrella, M., Maurino, A., & Ruffini, C. (2007). A FRAMEWORK AND A METHODOLOGY FOR DATA QUALITY ASSESSMENT AND MONITORING. Università di Milano Bicocca, Italy. Università di Milano Bicocca, Italy.

Chen, T. X., Meyer, M. D., Ganapathi, N., Liu, S. (., & Cirella, J. M. (2011). Improving Data Quality in Relational Databases: Overcoming Functional Entanglements. RTI International. RTI Press publication OP-0004-1105.

Even, A., & Shankaranarayanan, G. (2005). VALUE-DRIVEN DATA QUALITY ASSESSMENT. Boston University School of Management, IS Department, Boston. Boston: Boston University School of Management.

Sadiq, Z., & Duckham, M. (n.d.). STORING AND QUERYING SPATIALLY VARYING DATA QUALITY INFORMATION USING AN INTEGRATED SPATIAL RDBMS. The University of Melbourne, Victoria. Cooperative Research Centre for Spatial Information, Dept. of Geomatics.

Scannapieco, M. (2016). On the Meaningfulness of “Big Data Quality” (Invited Paper). Journal of Science and technology, 20.

Singh, R., & Singh, K. (2010). A Descriptive Classification of Causes of Data Quality Problems in Data Warehousing. IJCSI International Journal of Computer Science Issues, 7(3), 41-50.