Challenges for Visual Analytics

While visual analytics greatly benefits from the fast technological development, there are still some problems that remain to be solved. Many of them result from the specific applications of visual analytics and the requirements of the respective fields. Nevertheless, some challenges are common to many application areas.[1] They are outlined as follows.


One of the key challenges for visual analytics is scalability. It denotes the ability of the system to handle the continuously growing amount of data.[2] Scalability concerns both automated analysis and visualization. Visual analytics tools should be able to process extreme-scale heterogeneous data with multiple dimensions that is collected from different sources. This requires not only state of the art hardware with great computational power but also suitable software and algorithms that are capable of processing this information appropriately.[3] Moreover, the system needs to display these massive datasets effectively and in an interactive way. It should provide insight by creating a higher-level view of the data and at the same time be able to maximize the amount of detail when needed.[2]

Data issues

Data issues are another central challenge for visual analytics. The data preprocessing phase is a complex, time consuming, and costly process, especially when big data is involved.[4] Moreover, uncertainty and errors in the input data could cause incorrect interpretations and misleading analysis results.[5] There are various problems linked to the data used for visual analytics. Next, some of the major obstacles are outlined:[4]

  1. Data unavailability: in many cases there is missing data, e.g. because no information has been provided or because the data has been collected incorrectly.
  2. Access to data: it can be difficult to get access to data, especially in large companies.
  3. Data fragmentation: locating and integrating relevant data distributed across different databases is time consuming and also occurs often in large organizations.
  4. Data quality: data with missing values, data entered in an inconsistent way (e.g. date and time fields), data containing special characters, etc. make formatting a hard task to deal with and highlight the importance of standardization.
  5. Data shaping: users might need to modify the available data and create additional rows, columns, values, etc. in order to fit more data into their visualizations.
  6. Disconnect between data creation and data use: the data cannot be created in a way that will suit every eventual future use.
  7. Recordkeeping: records should be kept in a way that allows other users to understand the visual analytics process and its result without needing an additional explanation from the analysts.


Uncertainty (situation that involves imperfect or unknown information) is an issue for visual analytics that negatively impacts the entire process and could lead to incorrect results. Uncertainty can be related to the data (e.g. the way in which the data is collected and its quality), the model (e.g. inexperienced users without prior knowledge might find it hard to determine how many parameters to use or if the model is suitable for the purpose of the analysis), and the visualization (e.g. choosing inappropriate visualization technique or occurrence of visual artefacts caused by the resolution).[6] Users need to take uncertainties into account and be able to determine the quality of any stage of the process.[1] Furthermore, research suggest that novel techniques for uncertainty quantification and visualization will help users understand the risks and thus minimize the probability of drawing misleading conclusions.[7]

Human limitations

While the analytical power of computers increases rapidly, the human cognitive capability remains constant. Therefore, with their rather fixed abilities, the users are quickly becoming the bottleneck in visual analytics. It is a known fact that humans find it hard to visually detect relationships in a large amount of data and patterns often become white noise when the data reaches a certain size.[8] Another problem related to users is that they tend to get distracted by the ability to use visual analytics to explore their data and hence fail to keep track of the original purpose of the analysis.[9] The human also slows down the deployment of new techniques since many users refuse to change their working routines by using novel tools. Therefore, possible advantages remain unrealized and the full potential of visual analytics cannot be developed.[2] Last but not least, due to the complexity of the visual analytics process, it might be difficult for humans to evaluate it. All these issues make it one of the most challenging tasks for the field to come up with alternatives to compensate for human limitations.[7]

[1] Keim, D., Zang, L., Krstajić, M. and Simon, S. (2012) ‘Solving Problems with Visual Analytics: Challenges and Applications’, Journal of Multimedia Processing and Technologies, 3(1), pp. 1–11.
[2] Keim, D., Mansmann, F., Schneidewind, J. and Ziegler, H. (2006) ‘Challenges in Visual Data Analysis’, Tenth International Conference on Information Visualization. London, England, July 2006. pp. 9–16.
[3] Keim, D., Mansmann, F., Stoffel, A. and Ziegler, H. (2009) ‘Visual Analytics’, in Özsu, M.T. and Liu, L. (eds.) Encyclopedia of Database Systems. New York: Springer US, pp. 3341–3346.
[4] Lemieux, V.L., Gormly, B. and Rowledge, L. (2014) ‘Meeting Big Data challenges with visual analytics’, Records Management Journal, 24(2), pp. 122–141.
[5] Keim, D., Andrienko, G., Fekete, J.-D., Görg, C., Kohlhammer, J. and Melancon, G. (2008) ‘Visual Analytics: Definition, Process, and Challenges’, in Kerren, A., Stasko, J.T., Fekete, J.-D., and North, C. (eds.) Information Visualization. Berlin, Heidelberg: Springer-Verlag, pp. 154–175.
[6] Sacha, D., Senaratne, H., Kwon, B.C., Ellis, G. and Keim, D. (2016) ‘The Role of Uncertainty, Awareness, and Trust in Visual Analytics’, IEEE Transactions on Visualization and Computer Graphics, 22(1), pp. 240–249.
[7] Wong, P.C., Shen, H.-W., Johnson, C.R., Chen, C. and Ross, R.B. (2012) ‘The Top 10 Challenges in Extreme-Scale Visual Analytics’, IEEE Computer Graphics and Applications, 32(4), pp. 63–67.
[8] Wong, P.C., Shen, H.-W. and Chen, C. (2014) ‘Top Ten Interaction Challenges in Extreme-Scale Visual Analytics’, in Dill, J., Earnshaw, R., Kasik, D., Vince, J., and Wong, P.C. (eds.) Expanding the Frontiers of Visual Analytics and Visualization. United States: Springer, pp. 197–207.
[9] Lawton, G. (2009) ‘Users Take a Close Look at Visual Analytics’, IEEE Computer, 42(2), pp. 19–22.