Social Natural Language Processing Workshop
When: August 14 through 24 (Monday - Thursday, 2 weeks, 10am - 3pm) Where: New Computer Science, Stony Brook University Who: Integrating two audiences: Instructors: Lucie Flek (Professor, University of Bonn, Computer Science), Allie Lahnala (PhD Student, University of Bonn), Ryan Boyd (Senior Data Scientist, TikTok-ByteDance). Summary: In this 2-week workshop, participants will create interdisciplinary research projects that combine computational language processing methods such as that behind ChatGPT with social science research questions. Collaborating in teams, participants will design and implement research studies that analyze large-scale text data and gain insights into social behavior. Participants will gain hands-on experience in data collection, analysis, and interpretation, and will learn to communicate their findings to both technical and non-technical audiences. About the instructors: Lucie Flek is a renown researcher on machine learning applications in the field of Natural Language Processing (NLP), with a core expertise in the area of user modeling and stylistic variation. She investigates how individuals and sociodemographic groups differ in their language usage, and how this variation can be in return used in machine learning tasks to predict in-group behavior. Her broader expertise covers bias and mitigation techniques for the NLP field with respect to stereotype exaggeration, ethics issues, and the performance of machine learning models on underrepresented groups. She will be joined by Allie Lahnala, a PhD student studying computational social science and linguistics, particularly focused on supportive interactions in social media. Outside of research, she enjoys reading, playing video and board games, and spending time with her cat. Ryan Boyd is a renown computational social scientist, specializing in natural language analysis and the psychology of verbal behavior. He is the pimrary developer behind the software Linguistic Inquiry and Word Count and a generally curious person. Image Credit: Stable Diffusion 1.4, 2.1 (CompVis/stable-diffusion-v1-4, stabilityai/stable-diffusion-2-1-base) Detailed Description: Social Natural Language Processing is a field that combines social science research with computational methods and tools. This interdisciplinary field allows social scientists to conduct research on large-scale data sets, and use machine learning to identify patterns and relationships. Computational tools can be used to analyze large-scale text data, including social media posts, news articles, and historical documents. Text analysis methods include sentiment analysis, topic modeling, and machine learning-based classification, which can provide insights into the attitudes and beliefs of large groups of people. The proposed workshop will bring together two types of audiences: (1) CS AI students (2) Social and psychological researchers interested in language and AI Students of computer science will get to learn about the latest techniques specifically for language modeling and dialog, within the context of important social scientific research questions. Students of human-oriented fields, such as psychologists and sociologists, will collaborate with computer science students to explore a wide range of research topics related to language and communication. For example, automated emotion analysis can be useful in studying public opinion, user behavior, or political discourse. By examining patterns of language use, participants can gain insights into social norms, power dynamics, and cultural values. Psychologists and NLP researchers can collaborate to develop algorithms that analyze language use to identify personality traits, which can be useful in predicting behavior in certain contexts. Participants will learn about the various computational tools and methods used in social NLP, including sentiment analysis, topic modeling, and machine learning-based classification. They will also gain an understanding of how these methods can be used to analyze large-scale text data and gain insights into social norms, power dynamics, and cultural values. Participants will test the effectiveness of various NLP models on social science research questions. Organizers
Sponsors: |
