This document will cover the planning aspects for the first phase of this project, specifically the iLab project being undertaken by Perry Stephenson in Spring Semester 2016. The project will likely be handed back to Simon Knight following the completion of the iLab subject, and this project plan makes the assumption (below) that the project will be picked up again by another student in a future iLab.
This project (as posed by the client Simon Knight) is predominantly an exploratory analysis of one or more large academic writing datasets. As an entirely academic pursuit the project is focused entirely on insight discovery, and discussions with the client have not identified any formal requirements other than the implied requirements for high-quality analysis and documentation. The client has identified some potential outcomes from the analysis which will form a loose set of objectives:
- help researchers understand how the language used in academic contexts differs from common use
- flag opportunities for sections of text which should be reviewed to improve the quality of academic writing
- identify opportunities for collaboration between researchers working on similar topics
Whilst each of these objectives is entirely possible to achieve, it is unlikely that I will be able to achieve all of them in a single university semester. Additionally, these outcomes are not exhaustive and new objectives may be identified as the project progresses. This implies some key guiding principles for the project:
- the project must be managed and documented in such a way that future students can continue in future iLab semesters, ideally with minimal re-work
- finishing one thing well is more important than starting on many things
These principles give the client the best chance of benefiting from productive and fruitful iLab projects in future.
This iLab project will establish capability for the linguistic analysis of large text datasets. The key tasks which make up this phase of the project are:
- Identify and implement project structure and data science best practice to ensure project can be efficiently paused and handed over between iLab semesters
- Research and document the scope of modern text mining practices, and identify suitable algorithms and approaches to meet client objectives
- Additionally, identify what characteristics are important when sourcing additional text datasets for this project
- Implement and document proof-of-concept analysis to demonstrate the viability of specific approaches
- Build an interactive tool for demonstration
- Prepare and deliver presentation for client and stakeholders
This scope may be adjusted throughout the project if unforeseen delays make one or more tasks unachievable within the timeframes imposed by the academic calendar.
The schedule for the remainder of the project is hard to predict, as the critical path is entirely dependent on how many hours I spend on the project per week. It therefore seems more suitable to identify the key dates imposed by the university calendar, and the scope elements which need to be completed by that date.
| Date | Event | Scope Elements Delivered |
|---|---|---|
| 25 September | Assignment 1 - Blog Part B | 1, 2 |
| 13 October | Assignment 2 - In-class Presentations | 3, 4 |
| 22 October | Assignment 1 - Blog Part C | N/A |
| 6 November | Assignment 3 - Professional Showcase | 5 |
These deadlines may be adjusted as the formal requirements for each assessment are fleshed out in CIC Around discussions.
A formal quantified risk assessment is likely overkill for a project like this, however it is worth reflecting on some of the key risks for the project.
If I am unable to obtain the necessary skills in time to put them to use in this project, then there is a risk to scope items #3 and #4. Fortunately I already have a reasonable skillset in the text mining area, so I should be able to conduct detailed analysis and produce useful insights regardless of whether or not I am able to extend my skills further.
As I am working solo on this iLab and don't have any team members to keep me on track, there is a risk that I could lose focus and fail to achieve my objectives in line with the schedule above. I have partly mitigated this by selecting a topic that is closely related to my work, which results in some handy synergies and efficiencies. I have also chosen to take the iLab subject by itself this semester, ensuring that there are minimal distractions throughout the project. Personal time management is unlikely to be an issue.
Until the point at which I get the first useful insight from the data, there is a risk that there will be no significant findings. The insights could be obvious and/or insignificant, and there may be no practical use cases for what I develop. Some potential mitigations for this are:
- Accept that a null finding still has value, and adds to CIC's body of knowledge
- Ensure that I have the appropriate skills and toolset before commencing technical work on the project
- Make sure that the visualisation approach adds value, regardless of the analytical findings
None of these risks are deal-breakers, and they conveniently also have the same key risk control strategy: work really really hard. I'd better get to it!
The communication plan for this project is reasonably straight forward, as the client has taken responsibility for communications with all of the university stakeholders. These relationships do not significantly impact the project at this point, as I do not require their input to successfully execute each of the scope elements of my iLab project. Their involvement will become more important as the project progresses in future iLab semesters, which means that the output of my iLab will be a useful discussion aid when negotiating their further involvement. Given the current level of interest from university stakeholders it seems unlikely that any further communications will be required prior to the final showcase.
The final showcase will be a key communication with the university stakeholders, as it will effectively kick off future discussions about how this project might be of benefit to the university. For this reason, I will consider the stakeholder's interest in the project when designing my final showcase presentation.
It is challenging to develop a delivery plan at this stage as it is still unclear what is formally required as part of the assessment. I can however identify elements that I would like to present as part of the delivery:
- Static report detailing the analytical approaches undertaken, and showcasing the results
- Dynamic presentation (using a tool which I will build), exploiting the findings of the analysis to help academic authors understand more about the use of language in impact case studies
- Structured presentation as a showcase, potentially at UTS or the Data Science Sydney meetup group.
The allocation of these three elements into the formal assessment processes will be determined once the requirements are made clear.