Capturing Emotional Reactions to Data Visualization Preproposal

Background

As technology advances and more and more data is taken of nearly everything, there is an interesting gap in data about what people feel. This would be emotional data. Whether somebody is watching a movie, playing a videogame, or looking at a data visualization, people react emotionally. This sort of data can always be asked for afterwords by using a survey, but does that really give a good indication of what somebody was feeling at the exact moment they saw a scene in a horror film or a chart light up? A system that could accurately take emotional data would revolutionize the field of data science and could open up the doors to entirely new fields of research.

Executive Summary

This project should be able to take emotional data and make it available to a user. This can be done using body sensors or simply using facial recognition software. Currently, there is existing software that can recognize frowns, anger, happiness, etc through a webcam in real-time. Additionally, there are open-source tools for eye tracking. A tool that can combine these two data sources could be very powerful. The tool should be able to take data at specific time increments, or at predefined times, or could even hook up to programs to take data based on specific program events. Emotional data could change the landscape of data science, and a tool that makes taking data easy could be the first commercial tool to do so.

Viability

The main constraint to the success of this project is image quality. The tool mentioned above can recognize facial expressions in real time, so speed of this algorithm is not a problem. Additionally, open source eye tracking tools already exist. This project not only seems viable but it looks like it could be the next big step in data science. Image quality could be a huge problem though. Do we want this app to work on all sorts of cameras, or only certain ones? Should this work for mobile? These are questions that will have to be thought about more as the project develops and we become more comfortable with the tools that can be used.

Risks and Rewards

The primary risk of this project is privacy. Users will of course have to opt in to use the tool, but how do we protect user data from there? Massive companies that collect data consistently deal with these issues, how will we deal with it as well. Another risk is hoping that the APIs we plan on using work as desired. No software is perfect, and diving into a project that relies on third party software can be a great risk. One the other hand, this project has the potential to revolutionize the way companies collect and use data. The reward of this project is creating a tool that can and should be universally used.

Closing

Data is collected on everything possible. From what you hover your mouse over, to everything you type into a search bar. In many ways, this has made our lives significantly more convenient. Google knows what I want to search before I type it in. Amazon tells me what to buy. However, none of the data that is taken is based on human emotion, the thing that drives people. Imagine a non-evil company that was able to take emotional. This company would be able to find out what drives people. Emotional data is the future of data science, and this project can be one of the first steps in making it a reality.