MSc Computer Science - PROM05: Tools and Techniques

Project Management and Evaluation Tools

As I reach the end of Week 2 of the planning phase of my research project, this feels like a good point to reflect on how I plan to manage and evaluate the progress of my dissertation project. As my research focuses on developing a prototype deepfake detection tool for social media environments, effective project organisation and evaluation methods will be key to ensuring that both the technical and user-focused aspects of the work are addressed.

I have explored several project management and evaluation tools that could support the structure of my research. One approach that I am particularly interested in using is a combination of Agile-style task management and milestone planning. By breaking the project into smaller development stages, such as dataset preparation, model testing, interface design, and user evaluation, I will be able to monitor progress more clearly and identify potential risks at an early stage.

To support this, I plan to use tools like GitHub Projects to organise tasks, track development and issues, and document any technical changes. This platform allows progress to be visualised through boards and timelines, which is especially useful when managing multiple research activities alongside my academic deadlines. In addition to this, GitHub will provide my version control for code development, ensuring that experimental changes can be tested safely without compromising the prototype's stability.

In terms of evaluating project outcomes, I have begun to consider how both technical performance and user experience will be measured. From a technical perspective, I will utilise evaluation metrics such as detection accuracy, precision, recall, and processing time to assess how effectively the system identifies manipulated media. These metrics should help determine whether my working prototype performs reliably in realistic conditions, particularly on devices with limited processing power.

Alongside this, user-focused evaluation methods will also be important. I intend to explore the use of usability testing and structured questionnaires to gather feedback on how understandable and trustworthy the detection results appear from a user perspective. This will help me to assess whether the system supports informed decision-making rather than simply presenting technical outputs.

One challenge I have identified at this stage is balancing detailed evaluation with time constraints. Conducting extensive testing and user studies may be difficult to achieve within the available timeframe. To manage this, I plan to prioritise smaller-scale but more focused evaluations that can still provide meaningful insights into system performance and usability.

Overall, this week has highlighted the importance of structured project management and systematic evaluation within my research. By using appropriate tools and clearly defined assessment methods, I aim to ensure that the final prototype is not only technically effective but also practical and relevant for real-world social media environments.

References

Blischak, J.D., Davenport, E.R. and Wilson, G. (2016). A quick introduction to version control with Git and GitHub. PLOS Computational Biology. Available at: https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1004668 (Accessed 30 Jan. 26).

Sokolova, M. and Lapalme, G. (2009). A systematic analysis of performance measures for classification tasks. Information Processing & Management. Available at: https://www.sciencedirect.com/science/article/pii/S0306457309000259 (Accessed 30 Jan. 26).

Sommerville, I. (2016). Software Engineering. 10th edn. Harlow: Pearson Education.