The XAI Primer

We propose an exploratory interface depicting both XAI strategies and applications to support designers conceptualizing and developing new projects.

Explainable Artificial Intelligence (XAI) processes typically combine various explanation and verification strategies to support the analysis in different domains. Due to the increasing number of techniques and the variety of XAI methods deployed, deriving a comprehensive overview framework of different strategy combinations remains challenging.
The paper presents a proposal for a digital ideation space in which designers of XAI processes can derive inspiration and investigate existing works. The XAI Primer is designed based on the metaphor of a museum. Users can explore the presented ideation space as if they were artists visiting an art gallery. We enable serendipitous and guided explorations, allowing them to investigate and probe the state-of-the-art as a source of inspiration.

The exploration modes panel allows for jumping across layers and guided tours; the collection panel enables visitors collecting items during the navigation; the minimap helps in orienting users in digitalspace; the search bar, accessible by typing ctrl+c enables experts users to look for specific items or authors.


Figure 1: the figure depicts a diagrammatic representation of the components of the interface. Exploration modes allow visitors to dig into thhe space and reveal new layers of information.

The XAI Primer currently includes a collection of 71 Explainable Artificial Intelligence projects gathered from theVISxAI Workshop, Distill, Google Arts Experiments, IEEVIS Art Program, as well as independent artworks from the fields of art, media studies, communication design and data activism. We deliberately decided to focus on aselection of XAI projects from various research fields to open up a discussion of different approaches to XAI, including communication and explanation strategies across different disciplines.


Figure 2: the design process we followed from the data collection to the implementation and study.

Projects have been classified according to the Building Blocks Framework, previously defined by El-Assady et al. [1] in 2019 and tailored to the XAI Primer by Gobbo et El-Assady al. [2] in 2021. Items have been positioned in the layered space, then encoded as gliphs and enriched with additional materials. Guided tours have been defined by critically reading items' positioning and relationship. Before conducting the final qualitative study the system has been implementd as an interactive application built on top of React.js and D3.js.

  1. Mennatallah El-Assady, Wolfgang Jentner, Rebecca Kehlbeck, Udo Schlegel, Rita Sevastjanova, Fabian Sperrle, Thilo Spinner, and Daniel Keim. 2019. Towards XAI : Structuring the Processes of Explanations.In Proceedings of ACM CHI Workshop on Human-Centered Machine Learning

  2. Beatrice Gobbo, Mennatallah El-Assady. 2021.The XAI Primer : A Digital Ideation Space for Explainable Artificial Intelligence Strategies In Proceedings of ACM CHI Workshop on Human-Centered Explainable Artificial Intelligence.

The XAI Primer 2021