The Importance of Domain Knowledge to Data Analytics

The Importance of Domain Knowledge to Data Analytics

According to Camm et al. (2023), analytics includes using analytical techniques to compute statistical data. The main types of analytics include predictive, diagnostic, prescriptive, and descriptive data analytics. Predictive data analytics includes identifying correlations, trends, and causation. Subsequently, prescriptive analytics includes random testing and optimization, while descriptive analytics includes answering questions on what, how many, where, and when. On the other hand, diagnostic analytics includes examining data to determine an event and its cause. Data engineers can use pipeline-based analytics to establish the cause-and-effect relationship from a data set. Pipeline-based analytics includes using a data pipeline to analyze data (Smith et al., 2018). The data pipeline process is vital to organizations as they develop microservices dependent on the same data sources across platforms and products (Abdullah et al., 2017). A pipeline establishes a workflow of regular automatic updates to databases, reports, and other systems that investigate, store, and analyze data. Therefore, pipeline-based analytics can be useful in ensuring a continuous flow of information updates with little effort and less time. However, the effectiveness of pipeline-based analytics requires making some adjustments to the analytics to deal with the challenges of the technique and weaknesses. For instance, pipeline-based analytics can be supported by the categorization of the data in the pipeline to distinguish between different data sets and improve data governance, thus improving the accuracy of the insights derived from pipeline-based analytics. One of the improvements that should be made in pipeline-based analytics is the diversification of the pipelines that should be used for different types of data and reasons for analytics. Diversification can help improve the ease of categorizing data in the pipeline, thus enhancing the effectiveness of pipeline-based analytics. The second improvement that should be made is creating governance procedures across the pipeline to prevent mismanagement in the pipeline process, which could lead to unstable systems and data that cannot be managed effectively. Governance procedures can also facilitate the connection of data engineers involved in the pipeline process, thus promoting the analytics process.

References

Abdullah, A. S., Selvakumar, S., & Abirami, A. M. (2017). An introduction to data analytics. Advances in Business Information Systems and Analytics, 1–14. https://doi.org/10.4018/978-1-5225-2031-3.ch001

Camm, J. D., Cochran, J. J., Fry, M. J., & Ohlmann, J. W. (2023). Business analytics. South-Western.

Smith, M., Cronjaeger, S., Ershad, N., Nickle, R., & Peussner, M. (2018). Pipeline Data Analytics: Enhanced Corrosion Growth Assessment Through Machine Learning. Volume 1: Pipeline and Facilities Integrity. https://doi.org/10.1115/ipc2018-78364

ORDER A PLAGIARISM-FREE PAPER HERE

We’ll write everything from scratch

Domain Knowledge

Discuss at least two improvements that can be made in pipeline-based analytics. Explain your reasoning.