Tools & Methods for Understanding Scientific Collaborations
I’m interested in the design of tools and methodologies for understanding scientific and workplace collaborations. To this end, I’ve worked with members of the HDS Lab at UW to analyze scientific chat logs from a group of astrophysicists collecting data on supernovae. We analyzed some half a million chat messages collected from labs in the United States and France. We designed custom tools to code the affect in the messages, and to facilitate an understanding of relationships between affect and creative events in the workplace. We also worked to automate the process using machine learning approaches.
We went beyond a standard bag-of-words feature set in our analysis by using non-verbal cues, such as grammatical markers including unusual spelling (e.g., WHHHHAAAT!) and emoticons, and meta-information about the chat messages such as duration, length, and rate. We found that these decisions led to better classification results and supported our interpretation of affect as a dynamic phenomena. Our resulting pipeline for coding affect is available on GITHub as the command-line tool, ALOE.
My Role: I assisted in the coding and analysis of the chat logs and the design mock-ups for the collaborative chat tool ALOE.
Beyond the Individual: The Dynamic Features of Distributed Affect. Scott, T.J., Perry, D., Williams, A., Aragon, C.R., Proceedings of the ACM International Conference on Supporting Group Work, GROUP ’16, Sanibel Island, FL (2016). PDF
Statistical Affect Detection in Collaborative Chat. M. Brooks, K. Kuksenok, M. K. Torkildson, D. Perry, J. J. Robinson, T. J. Scott, O. Anicello, A. Zukowski, P. Harris, C. Aragon. CSCW 2013. PDF
Adapting Grounded Theory to Construct a Taxonomy of Affect in Collaborative Online Chat. T. J. Scott, K. Kuksenok, D. Perry, M. Brooks, O. Anicello, C. Aragon. SIGDOC 2012. PDF
Automating Large-Scale Annotation for Analysis of Social Media Content. K. Kuksenok, M. Brooks, J. J. Robinson, D. Perry, M. K. Torkildson, C. Aragon. Poster at 2nd Workshop on Interactive Visual Text Analytics, IEEE VisWeek (2012). PDF