Introduction
Data visualization is an efficient approach that significantly improves the interactivity of the examined data. In the talk “Next in Data Visualization,” five speakers discuss the innovative solutions about how to “explore” and “explain” information. The presentations primarily concern the ideas that might change the field in the future and provide new opportunities for professionals in various spheres. Ultimately, the current summarizing paper presents an overview of the crucial concepts from the “Next in Data Visualization” talk.
Allysa Goodman and Michelle Borkin
In the pre-presentation, Goodman emphasizes the importance of the “interactivity” concept – a focal point of data visualization. It implies that the primary objective of the field is to improve the visual accessibility of information and make life easier for professionals and regular consumers (Borkin, 2019). Consequently, Michelle Borkin supports this position and explains the importance of visualization implementation, specifically for healthcare applications. She emphasizes two primary concepts – network science and graph theory (Borkin, 2019). Both terms refer to the underlying framework of using graphs to represent information and the relationships between data nodes. This approach allows identifying the problematic areas by following the graph paths. Borkin (2019) implemented this method to significantly help doctors with heart disease, stroke, glaucoma, and brain tumor diagnostics. Moreover, Borkin and her team (2019) evaluated the differences between 2D and 3D visualization approaches and discussed the importance of method integration to combine the benefits of various frameworks. Ultimately, network science, graph theory, and the implementation of data visualization techniques in one image are the primary concepts of Borkin’s talk.
Arvind Satyanarayan
Arvind Satyanarayan starts his talk by discussing the technological advancement in machine learning, artificial intelligence (AI), and their potential to change people’s lives. Nevertheless, despite the increasing complexity, he mentions that interactivity is critical in data visualization (Satyanarayan, 2019). Therefore, he introduces the concept of declarative specification – an idea that visualization patterns are more significant than computation methods (Satyanarayan, 2019). Satyanarayan (2019) emphasizes how illustrative various approaches might be, including cross-filtering, interactive re-normalization, and other similar methods. He continues to discuss feature visualization – a concept of neural networking that identifies image characteristics and develops them through hidden layers (Satyanarayan, 2019). This idea is critical since it allows utilizing machine learning to distinguish designs that even humans cannot differentiate. Ultimately, these methods can be used for innovative solutions in data visualization in various fields.
Danielle Albers Szafir
Danielle Albers Szafir examines the relationship between cognitive processes and data science. She explains that a comprehensive understanding of human perception is essential to developing the most easy-to-read data visualization techniques (Szafir, 2019). One of the discussed concepts is separability – the idea that ensures that none of the visualization elements interfere with the interpretation of others (Szafir, 2019). Size, color, and shape are among the critical separability characteristics that provide notable visual signals in human perception. Moreover, Szafir (2019) explains that incorrect usage of visual elements in data graphs might lead to bias in outcomes, potentially resulting in an inaccurate perception of information. Therefore, it is vital to understand how to effectively visualize information using position, color, size, fillings, and distribution patterns to minimize the bias in outcomes. Lastly, Szafir (2019) discusses the possibility of manipulating visual elements to deceive the audience intentionally by addressing the peculiarities of human perception. Ultimately, the relationship between cognitive processes and data visualization is critical to maximizing the data’s interactivity.
Blacki Migliozzi
In the last part of the presentation, Migliozzi discusses how interactive visualization models might demonstrate the relationship between examined factors and influence decision-making. He reveals that innovative solutions in the field provide excellent opportunities for interdisciplinary collaboration on the example of climate change (Migliozzi, 2019). Moreover, it is possible to use interactive data visualization to raise awareness concerning relevant social and environmental issues since this approach delivers a powerful message to people (Migliozzi, 2019). Migliozzi (2019) further explains the connection between data visualization and human perception, stating that changing abstract statistics into easy-to-understand interactive graphs can make regular people understand the scale of climate change and global warming. Ultimately, the last part of the talk is dedicated to explaining the importance of data visualization through interactive patterns.
Personal Observations and Summary
In my opinion, all proposed data innovations have significant weight and might be used to improve the quality of the provided services in healthcare. For instance, the underlying concepts of cognitive processes and human perception are essential to developing easy-to-understand graphs and convincing people of the potential viability of the methods. However, I believe that the most critical data innovation concepts explained in the videos are network science and the collaborative approach. Borkin (2019) presented convincing evidence that visualizing diseases through a lens of graphs and nodes in 2D significantly improved the accuracy of diagnoses. I believe that it is absolutely crucial in healthcare since mistakes in the diagnostics process might lead to undesired consequences and harm patients. Moreover, they might cause financial damage to the organization due to the readmission requests and non-optimized working procedures. Ultimately, while all of the discussed data visualization techniques can be applied to healthcare, I believe that the concept of network science is the most critical one.
References
Borkin, M. [Harvard University]. (2019). Next in data visualization | Michelle Borkin | Radcliffe Institute [Video]. YouTube. Web.
Migliozzi, B. [Harvard University]. (2019). Next in data visualization | Blacki Migliozzi | Radcliffe Institute [Video]. YouTube. Web.
Satyanarayan, A. [Harvard University]. (2019). Next in data visualization | Arvind Satyanarayan | Radcliffe Institute [Video]. YouTube. Web.
Szafir, D. A. [Harvard University]. (2019). Next in data visualization | Danielle Albers Szafir | Radcliffe Institute [Video]. YouTube. Web.