Comparing Data Modelling Techniques:
In the age of ‘Big Data’, one where we are drowning in information from corporations, media, the world wide web etc… there has to be a way to structure this data in the information age. Companies are beginning to recognize that semantics is very important for the systems to communicate with each other and also with people who run them. A data model is then a drawing which represents data or things and the relationships between them.
A database is a collection of knowledge, which has been organised in some way so that it can be easily managed, accessed updated queried, retrieved etc… A relational database then is an organised structure of data which is used to leverage relationships and connections between objects. When modelling this form of databases one tends to use a relational schema, and the query this using SQL. For example the figure below shows an example of this database
This technique of modelling data is favorable when a user wants to specify data and certain queries, a user can input exactly what they want and leave it up to the software to bring back results which describe the data and the relationships between them.
Example of an relational database model:
The RDF or the Resource Description:
The RDF or the Resource Description Framework on the other hand is another type of data model. This RDF model is used as a method for conceptual modelling of information that is implemented in web resources. This type of data model uses a vary of syntax notations and data serialization formats and is based upon the idea of making statements about resources in the form of subject-predicate-object expressions. By using this type of data model it opens up data on a global scale and essentially enables anybody to refer to anything. The difference between the RDF model and an object-orientated model would be its use of object, subject, predicate as opposed to an entity, attribute, value approach. This, though, enables software to more easily exchange information throughout the web which then in turn results in the user gathering and receiving data from these databases more easily with greater efficiency and certainty. This as mentioned can then be done on a global scale which opens up the possibility for more users to access this material.
Example of an RDF model
Miller, Eric. Resource Description Framework (RDF) Model and Syntax. Accessed May 2017 https://www.w3.org/TR/WD-rdf-syntax-971002/
Pinczel, Balazs, Nagy, David, Kiss, Attila. ‘The Pros and Cons of RDF Structure Indexes’ Annales Univ. Sci. Budapest Vol 42. 2014. Pp 283-296
Schreibman Susan, Siemens Ray, Unsworth, John. A Companion to Digital Humanities Oxford: Blackwell, 2004.
Between Data and Knowledge (Data Visualisation)
Data originates from the latin word datum which roughly translates into “a thing which is given”. Data is raw and has no meaning it only just exists, it can be captured, modeled, inspected and then interpreted where it becomes information and knowing this information or having great experience with it then becomes knowledge. Before the computer age data would have been collected and captured only by individuals and then they would find conclusions or patterns in this data, leading this interpretation of the data and it becoming information and then knowledge. Now we are overwhelmed with different software and programmes that do the job of organising and analysing the data for us, and plotting this data on a graph of some sort. This opens up the opportunity of visually analysing and studying data which leads on to data visualisation.
Data visualisation is a powerful tool, not only does it make sense of the data by analysing it and transforming it into knowledge but also it is a ways of communication to make sense of said data and expressing that for the viewer. It creates patterns and trends in data where previously it was hard to make out. For example the raw data I decided to download from the CSO was mean temperatures recorded from different points in Ireland, I decided to focus on Malin Head (North), Roches Point(South), Dublin Airport (East) and Bellmullet (West). When reviewing the excel spreadsheet we can clearly see what the mean temperature was in a specific area within a specific month, we can also compare this figure to another month/location. However what this table fails to do is tell a story with these figures. This is where the data visualisation element comes into consideration. With the use of shape, colour position etc.. we can study the data a lot easier, while on the other hand it is communicated to us more clearly.
Above is the example of the spreadsheet. At a glance, the figures are all in a fairly close range to one another. However if we take a look at the data plotted out on a graph below the picture tells a different story, taking for example that in May the mean temperatures across the country were practically the same and that in December the temperature in Dublin was two degrees cooler than in the South. Above in the excel sheet the figures when presented in a table format couldn’t really truly express or communicate the story quite like the visual nature of the graph. I understand that the subject matter of the data I downloaded is quite trivial and there aren’t ground breaking results but I truly believe that without the visual component, one reviewing this data would not come to the same conclusions as one reviewing and analysing the table.(Image 2)
Malin Head (Yellow)
On the subject of the data visualisation programmes, as mentioned there are so many out there from very professional software, to your bog standard Microsoft Word. I began using Tableau which when starting off with one or two columns and rows worked perfectly however when I began adding on extra data, the programme began converting this to “null” (image 3) and as I’m sure happens to most when dealing with data, I lost patience and gave up. Instead I took the easy way out and googled the ‘top data visualisation tools’, where I came across what was quoted as being a tool which required “no technical abilities” called Data Hero. A tool which took only minutes to set up and was most importantly free was a user friendly one to use. I uploaded my CSV file, checked to make sure that the columns and rows, were all aligned and in correct order and pressed a button which then created a graph for me. I played around with it, the colours, different styles of charts etc.. and the end result was the graph pictured above (Image 2).
As I mentioned I haven’t made any groundbreaking results from this data exercise. But I hope I have conveyed the importance of data visualisation and the part it plays when coming to finding and conclusions and eventually information and knowledge for studies.
BonJour, Laurence, 1985. The Structure of empirical Knowledge. Cambridge, MA: Harvard University Press.
Few, Stephen. ‘Data Visualisation for Human Perception.’ The Interactive Design Foundation. Retrieved on 17th February 2016 https://www.interaction-design.org/literature/book/the-encyclopedia-of-human-computer-interaction-2nd-ed/data-visualization-for-human-perception
Central Statistics Office, retrieved on 17th February 2016 http://www.cso.ie/en/index.html