Ph.D. Research Reception

Presentations

Beyond Visuals: The Critical Impact of Text on Data Interpretation

Chase Stokes

While information visualizations leverage visual encodings of data to communicate key insights, they also rely heavily on text to situate and guide the reader. Through a series of crowdsourced studies, we examined how textual elements within visualizations, such as titles and annotations, influence readers' interpretations, preferences, predictions, and perceptions of author bias. We identify key trade-offs and impacts of integrating visual and textual information. Text on a visualization has a nuanced but significant impact on reader conclusions, a strong effect on preferences and perceptions of bias, and a relatively weak impact on predictions about future states of the data. Collectively, these studies highlight the roles of textual elements in visualization design, offering new insights into combining the two modes of information for data communication.

Once More, with Feeling: Analyzing Paralinguistic Performance in Film

Naitian Zhou

Translating from the script to an embodied performance requires numerous decisions on the part of the actor. In this project, we apply natural language and speech processing methods to hundreds of hours of movies to analyze the choices actors make in going from the written word to their spoken performance.

Expanding Perspectives on Data Privacy: Insights from Rural Togo

Zoe Kahn

Passively collected "big'' data sources are increasingly used to inform critical development policy decisions in low- and middle-income countries. While prior work highlights how such approaches may reveal sensitive information, enable surveillance, and centralize power, less is known about the corresponding privacy concerns, hopes, and fears of the people directly impacted by these policies --- people sometimes referred to as experiential experts. To understand the perspectives of experiential experts, we conducted semi-structured interviews with 124 people living in rural villages in Togo, shortly after an entirely digital cash transfer program was launched that used machine learning and mobile phone metadata to determine program eligibility. This talk begins by documenting participants' privacy concerns surrounding the introduction of big data approaches in development policy. We find that the privacy concerns of our experiential experts differ from those raised by privacy and development domain experts. To facilitate a more robust and constructive account of privacy, we discuss implications for policies and designs that take seriously the privacy concerns raised by both experiential experts and domain experts.

Dialogue Understanding between NLP and Cultural Studies

Kent Chang

Movies and TV series offer a wealth of conversational and social interactions that present complex pragmatic and sociolinguistic phenomena that warrant study. This line of work, which sits at the intersection of natural language processing (NLP) and cultural studies, often involves characterizing such interactions across a long temporal spa, which entails building algorithmic measuring devices that engage both deep learning and critical theory writ large.

In this talk, I will discuss my work on dialogue understanding in natural language processing, focusing on my paper, “Dramatic Conversation Disentanglement” (published in Findings of the Association for Computational Linguistics: ACL 2023): In this work, we draw on theoretical research in sociolinguistics, sociology, and film studies to operationalize a conversational thread (including the notion of a floor change) in dramatic texts, and use that definition to annotate a dataset of 10,033 dialogue turns (comprising 2,209 threads) from 831 movies. We compare the performance of several disentanglement models on this dramatic dataset, and apply the best-performing model to disentangle 808 movies. We see that, contrary to expectation, average thread lengths do not decrease significantly over the past 40 years, and characters portrayed by actors who are women, while underrepresented, initiate more new conversational threads relative to their speaking time.

How Users Perceive Fairness in Tipping in Online Food Delivery

Andrew Chong

Consumers increasingly interact with workers through technology-mediated marketplaces (TMMs)—environments where third-party companies manage interactions, control information, and constrain behavioral choices. In this project, we conducted interviews with 25 users of third-party food delivery services to explore why consumers might perceive tipping differently in TMMs. I discuss aspects of TMMs that complicate and shape consumer perceptions of a fair tip: 1) codification and its effect on the social meaning of tipping, 2) awareness of TMM companies as for-profit, third-party mediators, and 3) difficulties in assessing accountability. I also discuss how TMMs can affect how consumers relate to workers and influence efforts to build consumer-worker solidarity.

Designing for Care and Reflecting on Harm: Participatory Methods with Youth for Building Reflection and Introspection

Liza Gak

While design methods have fostered collaborating with youth around education and safety, there is also potential for youth to participate in design research to build reflection about their interpersonal relationships. As youth experience both connection and harm in online and offline spaces, it is important that youth build awareness around the roles of technologies in their close relationships. This presentation will feature two projects -- a qualitative study with teens about their location-sharing practices, and a series of design workshops where high school students wrote fictional stories about future technologies. Because digital technologies are so ubiquitous and embedded in children's daily lives, we see design methods as a useful, concrete method for building reflection, especially as technology increasingly mediates close relationships. We found that the ways that our participants reflected on the role of technology in their daily lives was intimately connected to their considerations of their interpersonal relationships.