Creating datasets for ML is an inherently human endeavor, as the data's heterogeneity mandates human intervention. However, most data workflows being one-time and hardly transferable leads to a lack of standardization and reusability. There has been a push to impose more structure on the data work process, but little is known about the implicit or "tacit" knowledge of data workers, i.e., "know-how"s that is difficult to transfer to others. Identifying and formalizing this knowledge can help data work improve, leading it from current "exploration" to more systematic "engineering." We interviewed 19 ML practitioners in this study to find "why" they use "what" tacit knowledge. As a result, we identified the following themes: 1) data is context/situation dependent, 2) human workers are inseparable from data, and 3) models must be understood to build data. We finally discuss future systematic supports and research to convert what is implicit to explicit.

CHI 2023 LBW.

As the saying goes, "Garbage in, Garbage out." Data significantly impacts dataset quality and model performance. However, data is notorious for its human-centric nature, making it subjective and complex. Constructing the data for ML systems requires human interventions, which cannot be easily quantified or structured. Therefore, ML practitioners undergo an iterative process of trials and errors, ad hoc solutions, and heterogeneous methods, which calls for standardization and structured data work. In this work, we suggest human-centric propositions for structured data construction.

HCAI Workshop @NeurIPS 2022.

Voice-based Conversational Agents (VCA) have served as personal assistants that support individuals with special needs. Adolescents with Autism Spectrum Disorder (ASD) may also benefit from VCAs to deal with their everyday needs and challenges, ranging from self-care to social communications. In this study, we explored how VCAs could encourage adolescents with ASD in navigating various aspects of their daily lives through the two-week use of VCAs and a series of participatory design workshops. Our findings demonstrated that VCAs could be an engaging, empowering, emancipating tool that supports adolescents with ASD to address their needs, personalities, and expectations, such as promoting self-care skills, regulating negative emotions, and practicing conversational skills. We propose implications of using off-the-shelf technologies as a personal assistant to ASD users in Assistive Technology design. We suggest design implications for promoting positive opportunities while mitigating the remaining challenges of VCAs for adolescents with ASD.

CHI 2021.

Inha Cha, Sung-In Kim, Hwajung Hong, Heejeong Yoo, Youn-kyung Lim