Tiffany France

My project uses machine learning emotion classification processes to create emotionally representative data from Project Gutenberg novels in the public domain. The data is fitted into a visualization that is rendered into book covers for the analyzed pieces. Both proprietary and handcrafted APIs are used to analyze emotions found in the corpuses of the 25 high-ranking books from Project Gutenberg. Multiple emotion recognition models are evaluated for accuracy and value in the literary space. The machine learning models are then compared to human evaluations of emotions, on the basis of existing psychoanalytic research. Results indicate that machine learning models can predict the emotions expressed in or evoked by novels and generate supporting book cover art. The final outcome of this project will be printable manuscripts with custom-generated covers. The visualizations are built using D3.js and data storytelling techniques. The charts will be imported into InDesign for final production and typesetting. Copyright issues are avoided by the use of books in the public domain, with original publication dates between 1800 and 1925. The book will contain the text by the original author as well as accompanying text and diagrams provided by this project. tiffanyfrance.com/ml-books