Generative AI (GenAI) is evolving from standalone tools to interconnected ecosystems that integrate chatbots, cloud platforms, and third-party services. While this ecosystem model enables personalization and extended services, it also introduces complex information flows and amplifies privacy risks. Existing solutions focus on system-level protections, offering little support for users to make meaningful privacy choices. To address this gap, we conducted two vignette-based survey studies with 486 participants and a follow-up interview study with 16 participants. We also explored users’ needs and preferences for privacy choice design across both GenAI personalization and data-sharing. Our results reveal paradoxical patterns: participants sometimes trusted third-party ecosystems more for personalization but perceived greater control in first-party ecosystems when data was shared externally. We discuss design implications for privacy choice interfaces that enhance transparency, control, and trust in GenAI ecosystems.

Roughly one in ten Americans move every year, bringing significant social and economic impact to both the places they move from and places they move to. We show that migration intent mined from internet search queries can forecast domestic migration and provide new insights beyond government data. We extract from a major search engine (Bing.com) 120 million raw queries with migration intent from 2014 to 2016, including origin and destination geographies, and the specific intent for migration such as whether the potential migration is housing or employment related. Using these queries, we map U.S. state level migration flows, validate them against government data, and demonstrate that adding search query-based metrics explains variance in migration prediction above robust baseline models. In addition, we show that the specific migration intent extracted from these queries unpack the differential demands of migrants with different demographic backgrounds and geographic interests. Examples include interactions between age, education, and income, and migration attributes such as buying versus renting housing and employment in technology versus manual labor job sectors. We discuss how local government, policy makers, and computational social scientists can benefit from this information.

Data visualizations in news articles (e.g., maps, line graphs, bar charts) greatly enrich the content of news articles and result in well-established improvements to reader comprehension. However, existing systems that generate news data visualiza-tions either require substantial manual effort or are limited to very specific types of data visualizations, thereby greatly re-stricting the number of news articles that can be enhanced. To address this issue, we define a new problem: given a news ar-ticle, retrieve relevant visualizations that already exist on the web. We show that this problem is tractable through a new system, VizByWiki, that mines contextually relevant data visualizations from Wikimedia Commons, the central file reposi-tory for Wikipedia. Using a novel ground truth dataset, we show that VizByWiki can successfully augment as many as 48% of popular online news articles with news visualizations. We also demonstrate that VizByWiki can automatically rank visualizations according to their usefulness with reasonable accuracy ([email protected] of 0.82). To facilitate further advances on our "news visualization retrieval problem", we release our ground truth dataset and make our system and its source code publicly available.

The widespread popularity of Pokémon GO presents the first opportunity to observe the geographic effects of location-based gaming at scale. This paper reports the results of a mixed methods study of the geography of Pokémon GO that includes a five-country field survey of 375 Pokémon GO players and a large scale geostatistical analysis of game elements. Focusing on the key geographic themes of places and movement, we find that the design of Pokémon GO reinforces existing geographically-linked biases (e.g. the game advantages urban areas and neighborhoods with smaller minority populations), that Pokémon GO may have instigated a relatively rare large-scale shift in global human mobility patterns, and that Pokémon GO has geographically-linked safety risks, but not those typically emphasized by the media. Our results point to geographic design implications for future systems in this space such as a means through which the geographic biases present in Pokémon GO may be counteracted.

In addition to encyclopedia articles and software, peer production communities produce structured data, e.g., Wikidata and OpenStreetMap's metadata. Structured data from peer production communities has become increasingly important due to its use by computational applications, such as CartoCSS, MapBox, and Wikipedia infoboxes. However, this structured data is usable by applications only if it follows standards. We did an interview study focused on OpenStreetMap's knowledge production processes to investigate how -- and how successfully -- this community creates and applies its data standards. Our study revealed a fundamental tension between the need to produce structured data in a standardized way and OpenStreetMap's tradition of contributor freedom. We extracted six themes that manifested this tension and three overarching concepts, correctness, community, and code, which help make sense of and synthesize the themes. We also offered suggestions for improving OpenStreetMap's knowledge production processes, including new data models, sociotechnical tools, and community practices (e.g. stronger leadership).

Catastrophic incidents associated with GPS devices and other personal navigation technologies are sufficiently common that these incidents have been given a colloquial nickname: "Death by GPS". While there is a significant body of work on the use of personal navigation technologies in everyday scenarios, no research has examined these technologies' roles in catastrophic incidents. In this paper, we seek to address this gap in the literature. Borrowing techniques from public health research and communication studies, we construct a corpus of 158 detailed news reports of unique catastrophic incidents associated with personal navigation technologies. We then identify key themes in these incidents and the roles that navigation technologies played in them, e.g. missing road characteristics data contributed to over 25% of these incidents. With the goal of reducing casualties associated with personal navigation technologies, we outline implications for design and research that emerge from our results, e.g. advancing "space usage rule" mapping, incorporating weather information in routing, and improving visual and audio instructions in complex situations.

Wikipedia-based studies and systems frequently assume that no two articles describe the same concept. However, in this paper, we show that this article-as-concept assumption is problematic due to editors' tendency to split articles into parent articles and sub-articles when articles get too long for readers (e.g. "Portland, Oregon" and "History of Portland, Oregon" in the English Wikipedia). In this paper, we present evidence that this issue can have significant impacts on Wikipedia-based studies and systems and introduce the sub-article matching problem. The goal of the sub-article matching problem is to automatically connect sub-articles to parent articles to help Wikipedia-based studies and systems retrieve complete information about a concept. We then describe the first system to address the sub-article matching problem. We show that, using a diverse feature set and standard machine learning techniques, our system can achieve good performance on most of our ground truth datasets, significantly outperforming baseline approaches.

Nearly any group is subject to turnover : some people leave, while others join. Turnover can be especially high in online groups, since participation typically is strictly voluntary. We investigated the effects of member turnover in online groups, specifically in Wikipedia's WikiProjects. We based our studies on theories from organizational science, which suggest that it is not just the amount of turnover, but the characteristics of those leaving and those joining that matter. We characterized leavers and newcomers by their prior productivity, tenure (in the group or community), and participation in other groups within the larger community. Furthermore, we considered the moderating effect of group size on turnover. We analyzed data from 88,427 editors who participated in 1,054 WikiProjects, finding that (1) the positive effects of newcomers to a group were larger than the negative effects of leavers, (2) prior productivity, tenure, and participation in other groups all played significant roles, and (3) the effects of leavers and newcomers were amplified in larger groups.

Wikipedia articles about places, OpenStreetMap features, and other forms of peer-produced content have become critical sources of geographic knowledge for humans and intelligent technologies. In this paper, we explore the effectiveness of the peer production model across the rural/urban divide, a divide that has been shown to be an important factor in many online social systems. We find that in both Wikipedia and OpenStreetMap, peer-produced content about rural areas is of systematically lower quality, is less likely to have been produced by contributors who focus on the local area, and is more likely to have been generated by automated software agents (i.e. "bots"). We then codify the systemic challenges inherent to characterizing rural phenomena through peer production and discuss potential solutions.