What is Data Visualisation?
Data visualisation is the study of representing data or information in a visual form. With the advancement of digital technologies, the scope of multimedia has increased manifold. Visuals in the form of graphs, images, diagrams, or animations have completely proliferated the media industry and the Internet.
It is an established fact that the human mind can comprehend information more easily if it is presented in the form of visuals. Instructional designers focus on abstract and model-based scientific visualisations to make the learning content more interesting and easy to understand.
Table of Content
- 1 What is Data Visualisation?
- 2 Data Visualization Techniques
- 3 Types of Data Visualisation
- 4 Applications of Data Visualisation
- 5 Visualising Big Data
News channels often integrate and present visuals related to accidents, natural disasters, weather reports, and survey results to incite a realistic imagination in the viewers’ mind. Nowadays, scientific data is also presented through digitally constructed images. These images are generally created with the help of computer software.
Visualisation is an excellent medium to analyse, comprehend, and share information. Let us see why:
- Visual images help to transmit a huge amount of information to the human brain at a glance.
- Visual images help in establishing relationships and distinctions between different patterns or processes easily.
- Visual interpretations help in exploring data from different angles, which help gain insights. Visualisation helps in identifying problems and understanding trends and outliers.
- Visualisations point out the key or interesting breakthroughs in a large dataset.
Data can be classified on the basis of the following three criteria irrespective of whether it is presented as data visualisation or infographics:
- Method of creation: It refers to the type of content used while creating any graphical representation.
- Quantity of data displayed: It refers to the amount of data that is represented. For example, geographical maps, companies financial data, etc.
- Degree of creativity applied: It refers to the extent to which the data is created graphically or designed in a colourful way or it is just showing some important data in black and white diagrams.
On the basis of the above evaluation, we can understand which is the correct form of representation for a given data type. Let’s discuss the various content types:
- Graph: A representation in which X and Y axes are used to depict the meaning of the information
- Diagram: A two-dimensional representation of information to show how something works
- Timeline: A representation of important events in a sequence with the help of self-explanatory visual material
- Template: A layout is a design for presenting information
- Checklist: A list of items for comparison and verification
- Flowchart: A representation of instructions that shows how something works or a step-by-step procedure to perform a task
- Mind map: A type of diagram which is used to visually organise information
Data Visualization Techniques
Data can be presented in various visual forms, which include simple line diagrams, bar graphs, tables, matrices, etc. Some data visualization techniques are as follows:
It is a 2D data representation of a curved line that moves constantly on the surface of a graph. The plotting of an isoline is based on the data arrangement rather than data visualisation. Figure 7.2 shows a set of isolines:
It is a 3D representation of an isoline. Isosurfaces are designed to represent points that are bound by a constant value in a volume of space, i.e. in a domain that covers 3D space.
Direct Volume Rendering (DVR)
It is a method used for obtaining a 2D projection for a 3D dataset. A 3D record is projected in a 2D form through DVR for a clearer and more transparent visualisation. Figure 7.4 shows a 2D DVR of a 3D image:
It is a field line that results from the velocity vector field description of the data flow. Figure 7.5 shows a set of streamlines:
It is a visual representation of locations within a specific area. It is depicted on a planar surface.
Parallel Coordinate Plot
It is a visualisation technique of representing multidimensional data. Figure 7.7 shows a parallel coordinate plot:
It is used to represent logical relations between finite collections of sets. Figure 7.8 shows a Venn diagram for a set of relations:
It is used to represent a chronological display of events. Figure 7.9 shows an example of a timeline for some critical event sets:
It is a representation of the relationships between sets. Figure 7.10 shows an example of a Euler diagram:
They represent graphs that are drawn using the hyperbolic geometry. Figure 7.11 shows a hyperbolic tree:
It represents a cluster such as a cluster of astronomic entities. Figure 7.12 shows a cluster diagram:
It is used to analyse various sets of multivariate objects. Figure 7.13 shows an ordinogram:
Types of Data Visualisation
You already know that data can be visualised in many ways, such as in the forms of 1D, 2D, or 3D structures. The following are the different types of data visualisations:
1D (Linear) data visualisation
In the linear data visualisation, data is presented in the form of lists. Hence, we cannot term it as visualisation. It is rather a data organisation technique. Therefore, no tool is required to visualise data in a linear manner.For example, a list of items organised in a predefined manner.
2D (Planar) data visualisation
This technique presents data in the form of images, diagrams, or charts on a plane surface. Cartogram and dot distribution map are examples of 2D data visualisation. Some tools used to create 2D data visualisation patterns are GeoCommons, Google Fusion Tables, Google Maps API, Polymaps, Tableau Public, etc. For example, choropleth, cartogram, dot distribution map, and proportional symbol map.
3D (Volumetric) data visualisation
In this method, data presentation involves exactly three dimensions to show simulations, surface and volume rendering, etc. Generally, it is used in scientific studies. Today, many organisations use 3D computer modelling and volume rendering in advertisements to provide users a better feel of their products. To create 3D visualisations, we use some visualisation tools that involve AC3D, AutoQ3D, TrueSpace, etc. For example, 3D computer models, surface rendering, volume rendering, and computer simulations.
Temporal data visualisation
Sometimes, visualisations are time dependent. To visualise the dependence of analyses on time, the temporal data visualisation is used, which includes Gantt chart, time series, sanky diagram, etc. TimeFlow, Timeline JS, Excel, Timeplot, TimeSearcher, Google Charts, Tableau Public, and Google Fusion Tables, etc. are some tools used to create temporal data visualisation.
Multidimensional data visualisation
In this type of data visualisation, numerous dimensions are used to present data. We have pie charts, histograms, bar charts, etc. to exemplify multidimensional data visualisation. Many Eyes, Google Charts, Tableau Public, etc. are some tools used to create multidimensional data visualisation.
Tree/Hierarchical data visualisation
Sometimes data relationships need to be shown in the form of hierarchies. To represent such kind of relationships, we use tree or hierarchical data visualisations. Examples of tree/hierarchical data visualisation include hyperbolic tree, wedge-stack graph, etc. Some tools to create hierarchical data visualisation are d3, Google Charts, and Network Workbench/Sci2.
Network data visualisation
It is used to represent data relations that are too complex to be represented in the form of hierarchies. Some examples of network data visualisation tools include matrix, node link diagram, hive plot. Pajek, Gephi, NodeXL, VOSviewer, UCINET, GUESS, Network Workbench/Sci2, sigma.js, d3/Protovis, Many Eyes, Google Fusion Tables, etc.
Applications of Data Visualisation
Data visualisation tools and techniques are used in various applications. Some of the areas in which we apply data visualisation are as follows:
Visualisation is applied to teach a topic that requires simulation or modelling of any object or process. Have you ever wondered how difficult it would be to explain any organ or organ system without any visuals? Organ system or structure of an atom is best described with the help of diagrams or animations.
Visualisation is applied to transform abstract data into visual forms for easy interpretation and further exploration.
Various applications are used to create 3D models of products for better viewing and manipulation. Real estate, communication, and automobile industry extensively use 3D advertisements to provide a better look and feel to their products.
Every field of science, including fluid dynamics, astrophysics, or medicine use visual representation of information. Isosurfaces and direct volume rendering are typically used to explain scientific concepts.
Systems visualisation is a relatively new concept that integrates visual techniques to better describe complex systems.
Multimedia and entertainment industry use visuals to communicate their ideas and information.
It refers to the science of analytical reasoning supported by the interactive visual interface. The data generated by social media interaction is interpreted using visual analytics techniques.
Visualising Big Data
Visual analysis of data is not a new thing. For years, statisticians and analysts have been using visualisation tools and techniques to interpret and present the outcomes of their analyses.
Almost every organisation today is struggling to tackle the huge amount of data pouring in every day. Data visualisation is a great way to reduce the turn-around time consumed in interpreting Big Data. Traditional visualisation techniques are not efficient enough to capture or interpret the information that Big Data possesses.
For example, such techniques are not able to interpret videos, audios, and complex sentences. Apart from the type of data, the volume and the speed with which it is generating pose a great challenge. Most of the traditional analytics techniques are unable to cater to any of these problems.
Big Data comprises both structured as well as unstructured form of data collected from various sources. Heterogeneity of data sources, data streaming, and real-time data are also difficult to handle using traditional tools. Traditional tools are developed using relational models that work best on static interaction.
Big Data is highly dynamic in function and therefore most traditional tools are not able to generate quality results. The response time of traditional tools is quite high, making it unfit for quality interaction.
Deriving Business Solutions
The most common notation used for Big Data is 3Vs—volume, velocity, and variety. But the most exciting feature is the way in which value is filtered from the haystack of data. Big Data generated through social media sites is a valuable source of information to understand consumer sentiments and demographics.
Almost every company nowadays is working with Big Data and facing the following challenges:
- Most data is in unstructured form
- Data is not analysed in real-time
- The amount of data generated is huge
- There is a lack of efficient tools and techniques
Considering all these factors, IT companies are focusing more on the research and development of robust algorithms, software, and tools to analyse the data that is scattered in the Internet space. Tools such as Hadoop are providing state-of-the-art technology to store and process Big Data.
Analytical tools are now able to produce interpretations on smartphones and tablets. It is possible because of the advanced visual analytics that is enabling business owners and researchers to explore the data to find out trends and patterns.
Turning Data into Information
The most exciting part of any analytical study is to find useful information from a plethora of data. Visualisation facilitates the identification of patterns in the form of graphs or charts, which in turn helps to derive useful information. Data reduction and abstraction are generally followed during data mining to get valuable information.
Visual data mining also works on the same principle as simple data mining; however, it involves the integration of information visualisation and human-computer interaction. Visualisation of data produces cluttered images that are filtered with the help of clutter-reduction techniques. Uniform sampling and dimension reduction are two commonly used clutter-reduction techniques.
Visual data reduction process involves automated data analysis to measure density, outliers, and their differences. These measures are then used as quality metrics to evaluate data-reduction activity. Visual quality metrics can be categorised as:
- Size metrics, (e.g. number of data points)
- Visual effectiveness metrics, (e.g. data density, collisions)
- Feature preservation metrics, (e.g. discovering and preserving data density differences)
In general, we can conclude that a visual analytics tool should be:
- Simple enough so that even non-technical users can operate it
- Interactive to connect with different sources of data
- Competent to create appropriate visuals for interpretations
- Able to interpret Big Data and share the information
Apart from representing data, a visualisation tool must be able to establish links between different data values, restore the missing data, and polish data for further analysis