Five steps of visual design

As you can see, information visualization? Can produce incredibly beautiful images and effectively convey information. The following figure is the communication organization chart of the now dissolved Enron Group.

Author/Copyright Owner: Kitware Inc Copyright Terms and License: CC BY-ND 2.0

Ben Schneiderman said? "The purpose of visualization is insight, not pictures. “

So, are there any visual design skills? Riccardo Mazar provided us with a five-step design method of information visualization in his book Introduction to Information Visualization. Let's have a look.

The whole design process is very simple. Once you review this process, it should be common sense:

1, define the problem

2. Define the data to be represented

3. Define the dimensions needed to represent the data.

4. Define the structure of data

5. Define the interaction required for visualization.

1. Definition problem

Like any user experience work; ? The first step is to define the problem to be solved in information visualization. This usually requires some user research? Answer questions; ? "What do my users need?" "How will they use it?"

You may try to explain something to users, or you may try to make them establish new contacts or observations; In order to clearly define the problem, you should also consider any specific factors unique to the user group, such as their education level or data processing ability. What kind of experience have they gained from past data? This will guide the complexity of the output and clarify the overall needs of users.

2. Define the data to be represented

Information visualization can represent three main types of data, and their mapping methods may be very different-so before you start designing, please clearly understand the data in your mind. What data will you use?

1, quantitative data? -This is digital data.

2. Ordered data? -Non-numeric data with internal order. (For example, think about every day of the week. )

3. confidential data? -There are neither numbers nor inherent sequential data. (such as business name or place name).

3. Define the dimensions needed to represent the data.

The number of dimensions or attributes of a data set must be carefully considered, because it will largely determine which data can be used for information visualization. The more dimensions represented in data, the more difficult it is to understand information visualization. Therefore, it is worth noting that data with a large number of dimensions may be more suitable for highly interactive presentation than static graphic presentation. According to the number of related dimensions to be studied, analysis can be divided into four types:

1, univariate analysis -Study the single dependent variable of independent variables.

2. Bivariate analysis? -Study two dependent variables of independent variables.

3. ternary analysis? -Study the three dependent variables of independent variables.

4. Multivariate analysis? -Study more than three dependent variables of independent variables.

?

Author/copyright owner: Chire. Copyright terms and license: CC BY-SA 3.0

Multivariate analysis images, in which the relationships between data points are many and related.

4. Define the structure of data

It's all about checking how data sets are related to each other. Common relationship structures include:

1, linear relationship? -Data can be displayed in a linear format, such as tables, vectors, etc.

2. Time relationship? -Data will change over time.

3. Spatial relationship? -Data related to the real world (such as map data or office floor plan) This is sometimes called geographical relationship.

4. hierarchical relationship? -Data related to location in the specified hierarchy (from office management structure to simple flow chart)

5. network relationship? -Data is related to other entities in the same data.

Zazie Hoko

Author/copyright owner: Nathanael Crawford. Copyright terms and license: CC BY-SA 3.0

An example of a hierarchical network model is shown above.

5. Define the required interaction from the visualization.

The last part of the design process requires you to understand the level of interaction required for user information visualization. There are three types of interactions:

1, static model? -These models are displayed "as is", such as the map in the road map set that you saved in the car. Users cannot modify them.

2. convertible model? -These models enable users to transform or modify data. They can allow users to change analysis parameters or choose different forms of visual mapping for data sets.

3. Operation mode? -Data are sometimes highly correlated, and it is a common interoperability method to manipulate some data graphs to obtain changes in related data, thus generating new insights. For example, we can quickly create graphic link operations through DataFocus.

Author/Copyright Owner: DataFocus Copyright Terms and License: CC BY-SA 3.0

? Tip: the process of designing information visualization may not be directly related to your final output, but it can make you decide which representation is most suitable for your users more wisely. By fully understanding the needs of users, the data they need to display, the relationships in the data and the model types, information visualization designers can provide visualization works that fully meet these needs.