Cody: An AI-based System to Semi-Automate Coding for Qualitative Research

What is it about?

With this paper, we introduce Cody - A system to semi-automate qualitative coding. Cody allows users to define code rules (like a search query for coding text instances) and picks up the user's existing codes to train a machine learning model to make code recommendations. We present results from two studies with qualitative researchers, in which we compare Cody with an established system for QDAS.

Featured Image

Photo by Mark Fletcher-Brown on Unsplash

Photo by Mark Fletcher-Brown on Unsplash

Why is it important?

With our paper, we conduct the first study of search-query style code rules that users themselves define in an intuitive coding prototype system. We combine these code rules with machine learning-based suggestions to support qualitative researchers during the time-consuming and tedious steps of qualitative coding. Also, we shed light on how researchers engage with explanations for code recommendations!