Tribe or Not? Critical Inspection of Group Differences Using TribalGram

Abstract

With the rise of AI and data mining techniques, group profiling and group-level analysis have been increasingly used in many domains, including policy making and direct marketing. In some cases, the statistics extracted from data may provide insights to a group’s shared characteristics; in others, the group-level analysis can lead to problems, including stereotyping and systematic oppression. How can analytic tools facilitate a more conscientious process in group analysis? In this work, we identify a set of accountable group analytics design guidelines to explicate the needs for group differentiation and preventing overgeneralization of a group. Following the design guidelines, we develop TribalGram , a visual analytic suite that leverages interpretable machine learning algorithms and visualization to offer inference assessment, model explanation, data corroboration, and sense-making. Through the interviews with domain experts, we showcase how our design and tools can bring a richer understanding of “groups” mined from the data.

VA + Embeddings STAR: A State‐of‐the‐Art Report on the Use of Embeddings in Visual Analytics

“…However, lines/links are often used as an augmentation approach to convey relationships between data entities, features, and other attributes. The word “alignment” is frequently used to describe the motivation for integrating embeddings [MWJ22, XWX * 22, AYL * 22], and computing clusters under different time snapshots or other constraints can inform changes or connections. To facilitate the interpretation of clustering results, parallel coordinate plots [XWX * 22, CGH * 22, AYL * 22], line charts [GZRP * 22], and node‐link diagrams [XO21, CGH * 22] are introduced as ways for users to find alignment for individual features, multidimensional attribute values, or within‐cluster and between‐cluster trends and patterns.…”

On Selective, Mutable and Dialogic XAI: a Review of What Users Say about Different Types of Interactive Explanations

“…Fig.1, TribalGram by Ahn et al[4] In the different explanations presented in the interface, users can see and compare the "blue camp" and the "red camp"' data groups, each represented respectively in blue or red. ' Fig.2C"Instance explanation", GAMUT by Hohman et al[47] …”