Applying Ontology-Guided Requirements Engineering to Improve Data Quality in Data Analytics

  • Porntida Kaewkamol

Student thesis: Phd

Abstract

The pervasiveness of data currently encourages the use of data analytics to support business and organisations. However, there are still challenges in executing data analytics in practice especially the quality issues of the datasets. Data analytics is generally an ad-hoc basis and is likely to have uncertain inputs and outcomes. This uncertainty might result in the unavailability of needed data, flawed decision, and project delays. This research proposes a framework as the artefacts to guide data scientists to identify and improve data quality at the early stages to minimise these uncertainties. The framework includes three artefacts 1) a supplementary process to guide data scientists to elicit, consider and improve data quality at the early phases. 2) a knowledge-based ontology for data quality requirements and improvement, and 3) an assistant tool to allow the interaction between the users and the system. After testing the framework by using a model of usefulness and ease of use, the results from the questionnaire showed users’ opinion and satisfactions towards the tool and the objective of the research.
Date of Award1 Aug 2023
Original languageEnglish
Awarding Institution
  • The University of Manchester
SupervisorPanagiotis Sarantopoulos (Supervisor) & Pedro Sampaio (Supervisor)

Keywords

  • Data quality
  • Requirements Engineering
  • Data analytics

Cite this

'