Capturing Emotional Reactions to Data Visualization
Background
Many corporations spend large amounts on creating data visualizations but there are no tools to track consumer’s emotional state when they are looking at a visualization. Data is being produced now at a large enough rate that many companies do not know how to handle the data. Because of this, companies are starting to do data visualization so the data is easier to digest than just looking at numbers. Also it can be easier to find trends visually than by just looking at numbers. A system to recognize emotion while looking at data visualization could help people to make decisions about how they visualize data and what reaction others might have to the data.
Executive Summary
There are multiple API’s that currently exist to track emotion in faces. Some of them are based on using Artificial Intelligence to learn about facial details and change over time while others use facial geometry to give percentages of emotions a face is representing. The system should be able to take in both pictures and videos of faces and then return some of the emotions present in the input. It could also give a emotional state over time. Lastly, it could also track eye movement to see what part of the data visualization the consumer is looking at.
Viability Analysis
The main constraints of this project will be to produce reliable and correct results for a person’s emotional state. The quality of video or pictures that are input into the program will dictate the reliability of the results. In addition, the different API’s may also produce different results so the most reliable one must be found. Also the API’s will have steep learning curves since they are based on Artificial Intelligence or Machine Learning. A large portion of time may also need to be devoted to train the program to recognize different emotions.
Risks and Rewards
There are privacy risks involved with this project because it involves using photos and videos of the people. Consent will need to be gained from every person who has their recording submitted to the application. There is also a risk that the API’s could be not reliable and not produce the correct results. The reward of this project would be an accurate tool that could be used by many different companies to help them improve their current data visualizations.
Closing
This project is one that could have a large upside potential. The field of emotional recognition is just taking off because of the strides in AI that have been made recently. There are not many tools that currently exist to do what the proposal is asking for, so completing this could give an advantage in the market for the company using it.
Print 
PDF