Data Modeling

Data modeling is all around us. In today’s society, data modeling is often overlooked. People do not realize the value of graphs and charts because we are constantly bombarded with them. Data modeling has allowed for the organization and structure of data around the world. Businesses use data models to relate data sets together and view them in bite-sized pieces. Adequate data modeling allows the content to be more digestible. As a result, it has become crucial when dealing with large data.

Representing data in an easy to understand way is key when dealing with important audiences. Data modeling allows this. For example, when presenting company finances or cost savings, it would be incredibly inefficient to show stakeholders large data tables. Data tables and raw data is hard to interpret by outside audiences. As a result, managers and employees represent data in graphs and charts. This allows stakeholders to see clear trends, patterns, and anomalies in the data. Could you imagine if the stock charts were data tables with second intervals? Understanding the stock’s path would be impossible, let alone comparing it with another stock. But, when we use graphs and charts, we can easily see whether the stock is rising long term and where dips occur. Investors can then use the chart and predict the dips.

Data models are also used in businesses to represent connections or correlations between datasets. Data models often have data structures, and most models look like a flipped tree, where the overarching “trunk” is at the top and branches to the other data structures. The data structures help businesses see where their data is coming from, and the fields it represents. Models also are heavily used in databases. Depending on the data stored and the amount, different data models dictate an efficient way to store the data. These models save companies large amounts of capital, as there is less unnecessary usage of resources.

With regards to data models themselves, they are split into 3 main stages: conceptual, logical, and physical data models. Conceptual models help define scope, as they show the data entities and their connections. The next step (logical model) defines the data entities’ requirements, with some models also defining the system to be implemented. Lastly, a physical data model provides the method of implementation of a system. It serves as a blueprint for a database and the flow of data.

pH2O Analytics Dataset 1

Personally, I use data modeling when working on pH2O Analytics. The goal of my non-profit is to spread educate about ocean acidification and its impacts. The data sets we use have hundreds of thousands of data points. We receive the data in the form of a large table. We then convert this table into a graph, and create predictive models based off the data. As a result, the data is visible, along with trends along the data. The public can understand the graphical data easily, and learn the gist of the data set. In conclusion, data modeling is a valuable resource, and as data sets continue to grow, the importance of data modeling will as well.

Image Source + Read More + pH2O Analytics