VIRTUAL: Natural Language Processing – 16 December @ 6:30PM

IEEE CC Zoom Event- Natural Language Processing – 16 December @ 6:30PM

Dr. Jessica Santana PhD UCSB Presents “Using Natural Language Processing to Measure Ethical Convergence in Scientific Discourse”

Professor Santana’s timely talk is sure to be very enlightening. When you Register for the event please include your email address so you can be included in our “Door Prize” drawing. Please use the link above to log onto the Zoom Event between 6:15 and 6:30 PM on December 16th. Meeting ID: 970 8235 5848, Passcode: 735956 

Using Natural Language Processing to Measure Ethical Convergence in Scientific Discourse

Natural language processing (NLP) holds many promises for understanding the complex relationship between scientific ethics and innovation. “Codes of ethics” emerge from scientific discourse. This study applies natural language processing and semantic network analysis to scientific discourse to discover how ethical norms become ethical codes. The goal of this research is to operationalize ethical codification in scientific discourse for the sociological study of boundary work in science and innovation. We do this by applying community detection to semantic and sociolect networks of scientific discourse and building predictive machine learning models for ethical sociolect community convergence using structured scientific discourse data (e.g. Web of Science) adapted to unstructured scientific discourse data (e.g. emails). In addition to scalable measurement of cultural norms in the production of science, the results of this research will also account for the dynamics of informal discourse captured in online social media streams, will connect linguistic variation with important social outcomes (e.g. innovation), and will reduce biases and data limitations of traditional scientific discourse analysis. Through this research, we ultimately aim to predict when the scientific community will label an innovation an ethical transgression or a scientific achievement. We demonstrate the application of this method in the context of the IEEE and ACM professional software engineering community.