Improved Decisions for Ocean FreightsAnalysis Learning
Expected Outcomes
There has been a recent surge in interests in Data Visualizations and their applications. As graduating students, we are looking for projects where we can learn, apply and export relevant and valuable skillsets out to the workforce at the end of the project. In the domain of Data Visualisations, we believe that there is a further need to differentiate between Dashboarding from Visualisations.
In the process of Visualisations, we expect to acquire skills such as learning to identify key indicators from a large dataset according to requirements; analyse that data and presenting it in charts, graphs that can offer insights to our client. Whereas with Dashboards as a product, while most of us would want to think of it as something the user would look at everyday for beautiful gathering of data; we believe it should amount to more than a mere “beautiful layout”, and more towards the needs of a user.
It is not an art lesson, nor is it a “high-level analysis”. It is afterall in its own right, a solution to a specific problem. And as such, more than to beautify or demonstrate visualisation technique, we would want to be able to identify the needs of our clients: what is it that he/she would want to see on opening the dashboard that would offer the answer to his/her questions. The collection of insights, ordered and structured, on each page would tell a story and offer not just another insight again; but, an answer to the question.
Collectively with the use of Visualisation and then Dashboarding, we would be able to learn not only the technology, the techniques but also the concepts of identifying indicators to producing visualisations and users’ needs in order to answer their questions; with which, we would be able to bring this skillsets to other problems as well.
Reflections
For this segment, we have classified it into 3 main points:
- Vision
- Understanding, to Analyze
- Wants vs. Needs
The key takeaway is Vision. Throughout the course of the project, generating out charts is often the fastest process. However, it is when they are to be put together where we face the greatest hurdle. A clear vision will allow us to decide on what insights to generate and reduce redundancy or wastage of time on a chart that turns out to be incoherent with the objectives.
Understanding, in order to analyze: data limitations could potentially lead to a wastage of time and resources put into generating analysis for the project. With proper comprehension of data such as how this data is generated, the potential issues that it might face; we can then estimate the resource required for analysis and for post-project replication.
Lastly, it is a race of what we want to achieve academically ersus the needs of the project. Academically, we tend to be interested in the deeper, more complex solutions of text mining, free input analysis etc. However, what we want to do, might not be what the project needs. Take for example, the mapping of codes and regions. We came up with and proved a methodology of using a dictionary based mapping for dirty data analysing. However, the methodology is long and the number of sheets required could impose a large amount of stress on replicating the work. There is a balance when it comes to work such as this versus that of an academic research, where we have to put the priorities of the client ahead. This is one situation that students often find themselves in and more often than never, students will choose the “wants” over the “needs”.
