Cardiff University | Prifysgol Caerdydd ORCA
Online Research @ Cardiff 
WelshClear Cookie - decide language by browser settings

A multi-disciplinary co-design approach to social media sensemaking with text mining

Rogers, David 2021. A multi-disciplinary co-design approach to social media sensemaking with text mining. PhD Thesis, Cardiff University.
Item availability restricted.

[thumbnail of PhD thesis] PDF (PhD thesis) - Accepted Post-Print Version
Available under License Creative Commons Attribution No Derivatives.

Download (7MB)
[thumbnail of Cardiff University Electronic Publication Form] PDF (Cardiff University Electronic Publication Form) - Supplemental Material
Restricted to Repository staff only

Download (794kB)


This thesis presents the development of a bespoke social media analytics platform called Sentinel using an event driven co-design approach. The performance and outputs of this system, along with its integration into the routine research methodology of its users, were used to evaluate how the application of an event driven co-design approach to system design improves the degree to which Social Web data can be converted into actionable intelligence, with respect to robustness, agility, and usability. The thesis includes a systematic review into the state-of-the-art technology that can support real-time text analysis of social media data, used to position the text analysis elements of the Sentinel Pipeline. This is followed by research chapters that focus on combinations of robustness, agility, and usability as themes, covering the iterative developments of the system through the event driven co-design lifecycle. Robustness and agility are covered during initial infrastructure design and early prototyping of bottom-up and top-down semantic enrichment. Robustness and usability are then considered during the development of the Semantic Search component of the Sentinel Platform, which exploits the semantic enrichment developed in the prototype, alpha, and beta systems. Finally, agility and usability are used whilst building upon the Semantic Search functionality to produce a data download functionality for rapidly collecting corpora for further qualitative research. These iterations are evaluated using a number of case studies that were undertaken in conjunction with a wider research programme, within the field of crime and security, that the Sentinel platform was designed to support. The findings from these case studies are used in the co-design process to inform how developments should evolve. As part of this research programme the Sentinel platform has supported the production of a number of research papers authored by stakeholders, highlighting the impact the system has had in the field of crime and security research

Item Type: Thesis (PhD)
Date Type: Completion
Status: Unpublished
Schools: Computer Science & Informatics
Subjects: Q Science > Q Science (General)
Date of First Compliant Deposit: 27 August 2021
Last Modified: 31 Aug 2021 13:40

Actions (repository staff only)

Edit Item Edit Item


Downloads per month over past year

View more statistics