On May 4, 2022, Sarah Morrison-Smith, Roman Family Teaching and Research Fellow of Computer Science, presented a paper at the ACM CHI Conference on Human Factors in Computing Systems. The paper, titled “QuAD: Deep-Learning Assisted Qualitative Data Analysis with Affinity Diagrams,” presents a prototype of Qualitative Affinity Diagrammer (QuAD), a novel affinity diagramming system. This paper was co-authored with several researchers, including Barnard students Alexandra Cheng ’22, Cindy Espinosa ’22, Ariel Goldman ’22, and Sabrina Meng ’23.
Morrison-Smith and her team argue that while affinity diagramming is an effective and efficient method for forming nuanced interpretations of qualitative data, the method does not scale well to large data sets. Instead, they propose QuAD, which successfully leverages computer-generated suggestions using deep learning to address the scalability of the diagramming process. The prototype of QuAD presented in the paper uses Bidirectional Encoder Representations from Transformers (BERT) and Girvan-Newman to generate grouping suggestions. The researchers aspire for their work to help others analyze large data sets in various contexts, including human-computer interaction.
The ACM CHI conference on Human Factors in Computing Systems is an international conference focused on Human-Computer Interaction (HCI). The conference brings together researchers and practitioners whose innovative projects use interactive digital technology to better society. You can watch a recording of Morrison-Smith and her team’s presentation here.