Abstract Dashboard guidance helps dashboard users better navigate interactive features, understand the underlying data, and assess insights they can potentially extract from dashboards. However, authoring dashboard guidance is a time consuming task, and embedding guidance into dashboards for effective delivery is difficult to realize. In this work, we contribute DashGuide, a framework and system to support the creation of interactive dashboard guidance with minimal authoring input. Given a dashboard and a communication goal, DashGuide captures a sequence of author‐performed interactions to generate guidance materials delivered as playable step‐by‐step overlays, a.k.a., dashboard tours. Authors can further edit and refine individual tour steps while receiving generative assistance. We also contribute findings from a formative assessment with 9 dashboard creators, which helped inform the design of DashGuide; and findings from an evaluation of DashGuide with 12 dashboard creators, suggesting it provides an improved authoring experience that balances efficiency, expressiveness, and creative freedom.

In our era of rapid technological advancement, the research landscape for writing assistants has become increasingly fragmented across various research communities. We seek to address this challenge by proposing a design space as a structured way to examine and explore the multidimensional space of intelligent and interactive writing assistants. Through community collaboration, we explore five aspects of writing assistants: task, user, technology, interaction, and ecosystem. Within each aspect, we define dimensions and codes by systematically reviewing 115 papers, while leveraging the expertise of researchers in various disciplines. Our design space aims to offer researchers and designers a practical tool to navigate, comprehend, and compare the various possibilities of writing assistants, and aid in the design of new writing assistants.

The use of Large Language Models (LLMs) for writing has sparked controversy both among readers and writers. On one hand, writers are concerned that LLMs will deprive them of agency and ownership, and readers are concerned about spending their time on text generated by soulless machines. On the other hand, AI-assistance can improve writing as long as writers can conform to publisher policies, and as long as readers can be assured that a text has been verified by a human. We argue that a system that captures the provenance of interaction with an LLM can help writers retain their agency, conform to policies, and communicate their use of AI to publishers and readers transparently. Thus we propose HaLLMark, a tool for visualizing the writer’s interaction with the LLM. We evaluated HaLLMark with 13 creative writers, and found that it helped them retain a sense of control and ownership of the text.

Large Language Models (LLMs) have created opportunities for designing chatbots that can support complex question-answering (QA) scenarios and improve news audience engagement. However, we still lack an understanding of what roles journalists and readers deem fit for such a chatbot in newsrooms. To address this gap, we first interviewed six journalists to understand how they answer questions from readers currently and how they want to use a QA chatbot for this purpose. To understand how readers want to interact with a QA chatbot, we then conducted an online experiment (N=124) where we asked each participant to read three news articles and ask questions to either the author(s) of the articles or a chatbot. By combining results from the studies, we present alignments and discrepancies between how journalists and readers want to use QA chatbots and propose a framework for designing effective QA chatbots in newsrooms.

Causal belief is a cognitive practice that humans apply everyday to reason about cause and effect relations between factors, phenomena, or events. Like optical illusions, humans are prone to drawing causal relations between events that are only coincidental (i.e., causal illusions). Researchers in domains such as cognitive psychology and healthcare often use logistically expensive experiments to understand causal beliefs and illusions. In this paper, we propose Belief Miner, a crowdsourcing method for evaluating people's causal beliefs and illusions. Our method uses the (dis)similarities between the causal relations collected from the crowds and experts to surface the causal beliefs and illusions. Through an iterative design process, we developed a web-based interface for collecting causal relations from a target population. We then conducted a crowdsourced experiment with 101 workers on Amazon Mechanical Turk and Prolific using this interface and analyzed the collected data with Belief Miner. We discovered a variety of causal beliefs and potential illusions, and we report the design implications for future research.

We presentaggregate query sculpting(AQS), a faceted visual query technique for large-scale multidimensional data. As a “born scalable” query technique, AQS starts visualization with a single visual mark representing an aggregation of the entire dataset. The user can then progressively explore the dataset through a sequence of operations abbreviated as P6:pivot(facet an aggregate based on an attribute),partition(lay out a facet in space), peek (see inside a subset using an aggregate visual representation), pile (merge two or more subsets), project (extracting a subset into a new substrate), andprune(discard an aggregate not currently of interest). We validate AQS with DATAOPSY, a prototype implementation of AQS that has been designed for fluid interaction on desktop and touch-based mobile devices. We demonstrate AQS and Dataopsy using two case studies and three application examples.

Many creative writing tasks (e.g., fiction writing) require authors to write complex narrative components (e.g., characterization, events, dialogue) over the course of a long story. Similarly, literary scholars need to manually annotate and interpret texts to understand such abstract components. In this paper, we explore how Natural Language Processing (NLP) and interactive visualization can help writers and scholars in such scenarios. To this end, we present Portrayal, an interactive visualization system for analyzing characters in a story. Portrayal extracts natural language indicators from a text to capture the characterization process and then visualizes the indicators in an interactive interface. We evaluated the system with 12 creative writers and scholars in a one-week-long qualitative study. Our findings suggest Portrayal helped writers revise their drafts and create dynamic characters and scenes. It helped scholars analyze characters without the need for any manual annotation, and design literary arguments with concrete evidence.

Sonification translates data into non-speech audio. Such auditory representations can make data visualization accessible to people who are blind or have low vision (BLV). This paper presents a sonification method for translating common data visualization into a blend of natural sounds. We hypothesize that people’s familiarity with sounds drawn from nature, such as birds singing in a forest, and their ability to listen to these sounds in parallel, will enable BLV users to perceive multiple data points being sonified at the same time. Informed by an extensive literature review and a preliminary study with 5 BLV participants, we designed an accessible data representation tool, Susurrus, that combines our sonification method with other accessibility features, such as keyboard interaction and text-to-speech feedback. Finally, we conducted a user study with 12 BLV participants and report the potential and application of natural sounds for sonification compared to existing sonification tools.

Data-centric AI has emerged as a new research area to systematically engineer the data to land AI models for real-world applications. As a core method for data-centric AI, data programming helps experts inject domain knowledge into data and label data at scale using carefully designed labeling functions (e.g., heuristic rules, logistics). Though data programming has shown great success in the NLP domain, it is challenging to program image data because of a) the challenge to describe images using visual vocabulary without human annotations and b) lacking efficient tools for data programming of images. We present Visual Concept Programming, a first-of-its-kind visual analytics approach of using visual concepts to program image data at scale while requiring a few human efforts. Our approach is built upon three unique components. It first uses a self-supervised learning approach to learn visual representation at the pixel level and extract a dictionary of visual concepts from images without using any human annotations. The visual concepts serve as building blocks of labeling functions for experts to inject their domain knowledge. We then design interactive visualizations to explore and understand visual concepts and compose labeling functions with concepts without writing code. Finally, with the composed labeling functions, users can label the image data at scale and use the labeled data to refine the pixel-wise visual representation and concept quality. We evaluate the learned pixel-wise visual representation for the downstream task of semantic segmentation to show the effectiveness and usefulness of our approach. In addition, we demonstrate how our approach tackles real-world problems of image retrieval for autonomous driving.

Implicit biases and stereotypes are often pervasive in different forms of creative writing such as novels, screenplays, and children’s books. To understand the kind of biases writers are concerned about and how they mitigate those in their writing, we conducted formative interviews with nine writers. The interviews suggested that despite a writer’s best interest, tracking and managing implicit biases such as a lack of agency, supporting or submissive roles, or harmful language for characters representing marginalized groups is challenging as the story becomes longer and complicated. Based on the interviews, we developed DramatVis Personae (DVP), a visual analytics tool that allows writers to assign social identities to characters, and evaluate how characters and different intersectional social identities are represented in the story. To evaluate DVP, we first conducted think-aloud sessions with three writers and found that DVP is easy-to-use, naturally integrates into the writing process, and could potentially help writers in several critical bias identification tasks. We then conducted a follow-up user study with 11 writers and found that participants could answer questions related to bias detection more efficiently using DVP in comparison to a simple text editor.

The widespread adoption of algorithmic decision-making systems has brought about the necessity to interpret the reasoning behind these decisions. The majority of these systems are complex black box models, and auxiliary models are often used to approximate and then explain their behavior. However, recent research suggests that such explanations are not overly accessible to lay users with no specific expertise in machine learning and this can lead to an incorrect interpretation of the underlying model. In this article, we show that a predictive and interactive model based on causality is inherently interpretable, does not require any auxiliary model, and allows both expert and non-expert users to understand the model comprehensively. To demonstrate our method we developed Outcome Explorer, a causality guided interactive interface, and evaluated it by conducting think-aloud sessions with three expert users and a user study with 18 non-expert users. All three expert users found our tool to be comprehensive in supporting their explanation needs while the non-expert users were able to understand the inner workings of a model easily.

Several prominent studies have shown that the imbalanced on-screen exposure of observable phenotypic traits like gender and skin-tone in movies, TV shows, live telecasts, and other visual media can reinforce gender and racial stereotypes in society. Researchers and human rights organizations alike have long been calling to make media producers more aware of such stereotypes. While awareness among media producers is growing, balancing the presence of different phenotypes in a video requires substantial manual effort and can typically only be done in the post-production phase. The task becomes even more challenging in the case of a live telecast where video producers must make instantaneous decisions with no post-production phase to refine or revert a decision. In this paper, we propose Screen-Balancer, an interactive tool that assists media producers in balancing the presence of different phenotypes in a live telecast. The design of Screen-Balancer is informed by a field study conducted in a professional live studio. Screen-Balancer analyzes the facial features of the actors to determine phenotypic traits using facial detection packages; it then facilitates real-time visual feedback for interactive moderation of gender and skin-tone distributions. To demonstrate the effectiveness of our approach, we conducted a user study with 20 participants and asked them to compose live telecasts from a set of video streams simulating different camera angles, and featuring several male and female actors with different skin-tones. The study revealed that the participants were able to reduce the difference of screen times of male and female actors by 43%, and that of light-skinned and dark-skinned actors by 44%, thus showing the promise and potential of using such a tool in commercial production systems.

Adaptation of the dataset shift has grown to be of great importance in machine learning problems in recent years. Reframing has emerged as a new machine learning technique that adapts the context changes between training and target domains. One of the advantages of reframing is that it can offer good performances with a limited amount of deployment data. Reframing has already been implemented in classification and regression by reusing labelled training data with the help of few labelled target data. However, reframing in clustering is still a challenging research problem because of its unsupervised nature. In this paper, we concentrate on building a reframing method for clustering. We also show the necessity and effectiveness of our method in contrast to retraining, which is the process of learning new model in the testing and deployment phases. Our evaluation results with extensive experiments using both synthetic and real-life datasets show that our method correctly identifies most of the shifts between datasets and builds better clustering model than retraining.