arXiv subjects

Ingmar Weber

Publications and source records attributed to Ingmar Weber.

At least 19 recordsLinked to original sources

arXiv2026

A global mobile network coverage raster product at 1km resolution, 1999--2030

Where a mobile signal is available shapes who can work, learn, bank, seek health care and respond to crises in the digital age, yet no globally consistent, sub-national record of mobile network coverage exists. We present such a record: annual 1km maps of the probability of 2G, 3G and 4G coverage for 214 countries and territories for the years 1999 to 2030. The maps are produced by three independent models: a calibrated machine-learning model, a techno-economic simulator of network build-out, and a spatial deep-learning model. The three estimates are then combined, per country and technology and in proportion to their measured accuracy, into a single best estimate with per-pixel 90% uncertainty bands; all four layers are released as part of the dataset. Because mobile roll-out closely follows a country's socio-economic conditions (population distribution, electrification, physical infrastructure), the models are grounded in existing geospatial data and tuned on 2,409 quality-screened operator-reported coverage maps, which are available up to 2020. For 2021--2024 the maps are predicted from recent geospatial data alone; for 2025--2030 they are extrapolated from demographic and infrastructure projections. On countries held out during training, the machine-learning model attains AUC 0.89--0.92. Baseline comparisons and the combined product's external validation are reported in Technical Validation. The dataset supports mapping the global digital divide, linking connectivity to household-survey outcomes, and humanitarian and infrastructure planning.

cs.CY↗

arXiv2026

Characterizing Bluesky Content Moderation Service: From Automation of Service to Landscape of Harms

Empirical research on content moderation is fundamentally constrained by the opaque deployment of moderation systems on major social media platforms. To this end, the recent emergence of decentralized platforms with transparent, public moderation logs presents an unprecedented opportunity for independent audits. In this work, we leverage this architectural transparency to conduct the first large-scale audit of the default moderation system on Bluesky, the Bluesky Moderation Service (BMS). Analyzing its 10.6M moderation labels from 2025, we investigate three foundational aspects: (i) its mechanism (the degree of automation versus human oversight), (ii) its efficacy (accuracy in detecting harms), and (iii) its purpose (the landscape of harms it identifies). Our findings reveal a human-AI collaborative system where labels for sexual and graphic content are applied automatically in seconds, while nuanced and high stakes labels require more human oversight, taking hours or days. Through a manual annotation study, we find the BMS operates with high precision (0.837), but struggles with low recall (0.222), with our annotators identifying 4.5$\times$ more harmful content than the moderation system in a random sample. Finally, unsupervised clustering of the most frequently applied labeled posts uncovers detected harms ranging from hostility in discourse toward protected groups to the spread of sexually explicit and other graphic content. Our work offers a look into the operational realities of a deployed moderation system, providing a concrete data-driven foundation for designing more effective and transparent moderation systems.

cs.CY↗

arXiv2026

Can Foundation Models Moderate Online Content? Evaluating Instruction- vs. Example-Driven Policy Operationalization

The growing complexity of content moderation policies presents a critical challenge for their consistent operationalization. While foundation models possess the basic capabilities needed to confront this challenge, whether they can reliably moderate online content remains an unanswered question. In this paper, we systematically compare two competing paradigms for Vision-Language Model (VLM) guidance: an instruction-driven approach where models reason from policy precepts, and an example-driven approach where they generalize from prior precedents. We ground this investigation in ModerationBench, a new benchmark of 4,000 manually annotated, in-the-wild posts from the Bluesky platform. Our experiments reveal that foundation models can substantially outperform Bluesky's deployed moderation system, nearly tripling its $F_1$ score (0.60 vs. 0.22) on Random Posts in the benchmark, with both instruction- and example-driven paradigms achieving comparable peak effectiveness. Our findings thus chart a path toward reliable and adaptable policy operationalization at scale.

cs.CL↗

arXiv2026

Divergent Paths to Depolarization: Dialogue Design Shapes the Intergroup Attitudinal Effects of AI-Assisted Political Argumentation

Structured argumentative dialogues where interlocutors deliberate on opposing political ideas are known to promote perspective-taking and reduce political polarization, but finding willing partners is difficult as Americans increasingly shun political discussions. AI dialogue partners offer a scalable framework for such open-mindedness exercises, but how the format of human-AI dialogues shapes their benefits remains unclear. This study seeks to fill the gap with a preregistered two-session online experiment with 527 US participants. As the primary experimental manipulation, participants were assigned to argue either for or against their pre-existing attitude on a contested political issue, engaging either with an AI chatbot or a solitary essay task. The AI conditions further varied in the chatbot's interaction style (adversarial or collaborative) and the presence of an additional financial incentive. The results show that attitude-congruent dialogues more strongly reduced polarization than attitude-incongruent dialogues immediately after the exchange. By contrast, an exploratory analysis suggested a delayed increase in cognitive empathy following attitude-incongruent dialogues, a pattern consistent with the account of sleeper effects. While the conversation style had little influence on the effects of attitude-congruent dialogues, a collaborative discussion tended to make attitude-incongruent dialogues more effective, narrowing the immediate effect gap. Additional financial incentives did not alter outcomes. Given the heterogeneity, the AI conditions were not universally more effective forming favorable intergroup attitudes in pooled comparisons between AI and non-AI conditions. The findings caution against a simplistic view of AI dialogues as a silver bullet for depolarization and highlight dialogue design as a key determinant of effective AI-mediated attitudinal interventions.

cs.CY↗

arXiv2026

No one likes it hot, but hotter cities adjust by staying active later

Extreme heat suppresses urban activity, but its effects need not be uniform across climates or across the day. Using data on activity at points of interest in 20 cities spanning temperate, tropical, and arid environments, we show that hot days reduce activity overall while shifting it away from midday and toward later hours. This rescheduling is substantially stronger in historically hotter cities, which exhibit smaller losses and larger evening substitution. To understand these changes, we introduce a Bactrian index of bimodality, which measures the degree to which a city's daily activity profile has one hump or two - one during the day and another during the evening. Arid desert cities like Doha, Amman, and Kuwait City are more Bactrian in level, but cities like Milan become Bactrian on hot days. Together, our results suggest that adaptation to heat in cities operates less through avoiding activity altogether than through moving it to cooler hours. This provides channels for adaptation in cooler cities, but it also suggests limits to adaptation in warmer ones: as evenings become warmer, these too may become intolerable.

physics.soc-ph↗

arXiv2026

Designing Safe and Accountable GenAI as a Learning Companion with Women Banned from Formal Education

In gender-restrictive and surveilled contexts, where access to formal education may be restricted for women, pursuing education involves safety and privacy risks. When women are excluded from schools and universities, they often turn to online self-learning and generative AI (GenAI) to pursue their educational and career aspirations. However, we know little about what safe and accountable GenAI support is required in the context of surveillance, household responsibilities, and the absence of learning communities. We present a remote participatory design study with 20 women in Afghanistan, informed by a recruitment survey (n = 140), examining how participants envision GenAI for learning and employability. Participants describe using GenAI less as an information source and more as an always-available peer, mentor, and source of career guidance that helps compensate for the absence of learning communities. At the same time, they emphasize that this companionship is constrained by privacy and surveillance risks, contextually unrealistic and culturally unsafe support, and direct-answer interactions that can undermine learning by creating an illusion of progress. Beyond eliciting requirements, envisioning the future with GenAI through participatory design was positively associated with significant increases in participants' aspirations (p=.01), perceived agency (p=.01), and perceived avenues (p=.03). These outcomes show that accountable and safe GenAI is not only about harm reduction but can also actively enable women to imagine and pursue viable learning and employment futures. Building on this, we translate participants' proposals into accountability-focused design directions that center on safety-first interaction and user control, context-grounded support under constrained resources, and offer pedagogically aligned assistance that supports genuine learning rather than quick answers.

cs.CY↗

arXiv2026

Teacher Professional Development on WhatsApp and LLMs: Early Lessons from Cameroon

AI in education is commonly delivered through web-based systems such as online forms and institutional platforms. However, these approaches can exclude teachers in low-resource contexts, where everyday mobile platforms like WhatsApp serve as primary digital infrastructure. To address this gap, we present a field pilot in Cameroon that deploys a WhatsApp-based chatbot with LLM-supported content for teacher professional development (TPD), compared with an online form baseline. The system was evaluated through a mixed-methods study with 47 primary school teachers, integrating quantitative measures with qualitative insights from interviews and participant feedback. Results show that the chatbot was rated higher in perceived usability and overall experience, while learnability remained comparable. These improvements were driven by platform familiarity, low interaction overhead, and the modular structure of LLM-supported content, but were constrained by connectivity limitations, prepaid data costs, and multilingual needs (English/French). Building on these findings, we outline design directions for multilingual, culturally grounded interaction and for supporting prompting and reflection in AI use. More broadly, this work points to Thoughtful AI that supports reflection, relevance, and sustained professional growth.

cs.HC↗

arXiv2026

You See It, They Don't: An Exploratory Study of User-to-User Variation in Instagram Comments

In March 2025, Meta announced a new AI system to rank the order of the comments shown to Instagram users. With existing research showing how feed personalization systems can lead to increased polarization, the introduction of this new system raises similar questions. This paper presents a small-scale exploratory study examining whether the ranking system produces systematic differences in visible comments shown to different users, particularly for news-related content. Using four sock-puppet accounts varying in gender and political leaning, we collect visible comments on posts from ten news and ten non-news accounts. This collection is repeated twice from two VPN locations to assess location effects. We ask 1) how many visible comments vary across different users, 2) is this variation higher for news accounts than non-news accounts, and 3) can user-attributes like gender, political leaning, and location systematically explain the observed variation. Contrary to our expectations, we find that visible comments on news posts are less likely to vary across users than those on non-news posts. Variation is better explained by account metrics like comment and follower counts than by user attributes. These findings provide an initial glimpse into personalized comment ranking on Instagram and motivate larger, more systematic audits of how comment personalization may shape online discourse. To support further research, we provide the code to collect comments and the data upon request.

cs.CY↗

arXiv2026

Bowling with ChatGPT: On the Evolving User Interactions with Conversational AI Systems

Recent studies have discussed how users are increasingly using conversational AI systems, powered by LLMs, for information seeking, decision support, and even emotional support. However, these macro-level observations offer limited insight into how the purpose of these interactions shifts over time, how users frame their interactions with the system, and how steering dynamics unfold in these human-AI interactions. To examine these evolving dynamics, we gathered and analyzed a unique dataset InVivoGPT: consisting of 825K ChatGPT interactions, donated by 300 users through their GDPR data rights. Our analyses reveal three key findings. First, participants increasingly turn to ChatGPT for a broader range of purposes, including substantial growth in sensitive domains such as health and mental health. Second, interactions become more socially framed: the system anthropomorphizes itself at rising rates, participants more frequently treat it as a companion, and personal data disclosure becomes both more common and more diverse. Third, conversational steering becomes more prominent, especially after the release of GPT-4o, with conversations where the participants followed a model-initiated suggestion quadrupling over the period of our dataset. Overall, our results show that conversational AI systems are shifting from functional tools to social partners, raising important questions about their design and governance.

cs.HC↗

arXiv2025

Safe for Whom? Rethinking How We Evaluate the Safety of LLMs for Real Users

Safety evaluations of large language models (LLMs) typically focus on universal risks like dangerous capabilities or undesirable propensities. However, millions use LLMs for personal advice on high-stakes topics like finance and health, where harms are context-dependent rather than universal. While frameworks like the OECD's AI classification recognize the need to assess individual risks, user-welfare safety evaluations remain underdeveloped. We argue that developing such evaluations is non-trivial due to fundamental questions about accounting for user context in evaluation design. In this exploratory study, we evaluated advice on finance and health from GPT-5, Claude Sonnet 4, and Gemini 2.5 Pro across user profiles of varying vulnerability. First, we demonstrate that evaluators must have access to rich user context: identical LLM responses were rated significantly safer by context-blind evaluators than by those aware of user circumstances, with safety scores for high-vulnerability users dropping from safe (5/7) to somewhat unsafe (3/7). One might assume this gap could be addressed by creating realistic user prompts containing key contextual information. However, our second study challenges this: we rerun the evaluation on prompts containing context users report they would disclose, finding no significant improvement. Our work establishes that effective user-welfare safety evaluation requires evaluators to assess responses against diverse user profiles, as realistic user context disclosure alone proves insufficient, particularly for vulnerable populations. By demonstrating a methodology for context-aware evaluation, this study provides both a starting point for such assessments and foundational evidence that evaluating individual welfare demands approaches distinct from existing universal-risk frameworks. We publish our code and dataset to aid future developments.

cs.AI↗

arXiv2025

Digital Diasporas: How Origin Characteristics and Host-Native Distance Shape Immigrants' Online Cultural Retention

Immigrants bring unique cultural backgrounds to their host countries. Subsequent interplay of cultures can lead to either a melting pot, where immigrants adopt the dominant culture of the host country, or a mosaic, where distinct cultural identities coexist. The existing literature primarily focuses on the acculturation of immigrants, specifically the melting pot hypothesis. In contrast, we attempt to identify the antecedents of the mosaic hypothesis or factors that enhance (or diminish) the propensity for cultural retention among immigrants. Based on Facebook advertising data for immigrants from 8 countries residing in the USA, our findings suggest that greater host-native distance is linked to higher online cultural retention, and while origin country context is statistically significant, its impact is generally smaller.

cs.HC↗

arXiv2025

Hope, Aspirations, and the Impact of LLMs on Female Programming Learners in Afghanistan

Designing impactful educational technologies in contexts of socio-political instability requires a nuanced understanding of educational aspirations. Currently, scalable metrics for measuring aspirations are limited. This study adapts, translates, and evaluates Snyder's Hope Scale as a metric for measuring aspirations among 136 women learning programming online during a period of systemic educational restrictions in Afghanistan. The adapted scale demonstrated good reliability (Cronbach's {\alpha} = 0.78) and participants rated it as understandable and relevant. While overall aspiration-related scores did not differ significantly by access to Large Language Models (LLMs), those with access reported marginally higher scores on the Avenues subscale (p = .056), suggesting broader perceived pathways to achieving educational aspirations. These findings support the use of the adapted scale as a metric for aspirations in contexts of socio-political instability. More broadly, the adapted scale can be used to evaluate the impact of aspiration-driven design of educational technologies.

cs.CY↗

arXiv2025

A Global Dataset of Location Data Integrity-Assessed Reforestation Efforts

Afforestation and reforestation are popular strategies for mitigating climate change by enhancing carbon sequestration. However, the effectiveness of these efforts is often self-reported by project developers, or certified through processes with limited external validation. This leads to concerns about data reliability and project integrity. In response to increasing scrutiny of voluntary carbon markets, this study presents a dataset on global afforestation and reforestation efforts compiled from primary (meta-)information and augmented with time-series satellite imagery and other secondary data. Our dataset covers 1,289,068 planting sites from 45,628 projects spanning 33 years. Since any remote sensing-based validation effort relies on the integrity of a planting site's geographic boundary, this dataset introduces a standardized assessment of the provided site-level location information, which we summarize in one easy-to-communicate key indicator: LDIS -- the Location Data Integrity Score. We find that approximately 79\% of the georeferenced planting sites monitored fail on at least 1 out of 10 LDIS indicators, while 15\% of the monitored projects lack machine-readable georeferenced data in the first place. In addition to enhancing accountability in the voluntary carbon market, the presented dataset also holds value as training data for e.g. computer vision-related tasks with millions of linked Sentinel-2 and Planetscope satellite images.

cs.LG↗

arXiv2025

VME: A Satellite Imagery Dataset and Benchmark for Detecting Vehicles in the Middle East and Beyond

Detecting vehicles in satellite images is crucial for traffic management, urban planning, and disaster response. However, current models struggle with real-world diversity, particularly across different regions. This challenge is amplified by geographic bias in existing datasets, which often focus on specific areas and overlook regions like the Middle East. To address this gap, we present the Vehicles in the Middle East (VME) dataset, designed explicitly for vehicle detection in high-resolution satellite images from Middle Eastern countries. Sourced from Maxar, the VME dataset spans 54 cities across 12 countries, comprising over 4,000 image tiles and more than 100,000 vehicles, annotated using both manual and semi-automated methods. Additionally, we introduce the largest benchmark dataset for Car Detection in Satellite Imagery (CDSI), combining images from multiple sources to enhance global car detection. Our experiments demonstrate that models trained on existing datasets perform poorly on Middle Eastern images, while the VME dataset significantly improves detection accuracy in this region. Moreover, state-of-the-art models trained on CDSI achieve substantial improvements in global car detection.

cs.CV↗

arXiv2025

Misleading through Inconsistency: A Benchmark for Political Inconsistencies Detection

Inconsistent political statements represent a form of misinformation. They erode public trust and pose challenges to accountability, when left unnoticed. Detecting inconsistencies automatically could support journalists in asking clarification questions, thereby helping to keep politicians accountable. We propose the Inconsistency detection task and develop a scale of inconsistency types to prompt NLP-research in this direction. To provide a resource for detecting inconsistencies in a political domain, we present a dataset of 698 human-annotated pairs of political statements with explanations of the annotators' reasoning for 237 samples. The statements mainly come from voting assistant platforms such as Wahl-O-Mat in Germany and Smartvote in Switzerland, reflecting real-world political issues. We benchmark Large Language Models (LLMs) on our dataset and show that in general, they are as good as humans at detecting inconsistencies, and might be even better than individual humans at predicting the crowd-annotated ground-truth. However, when it comes to identifying fine-grained inconsistency types, none of the model have reached the upper bound of performance (due to natural labeling variation), thus leaving room for improvement. We make our dataset and code publicly available.

cs.CL↗

arXiv2025

Learn, Explore and Reflect by Chatting: Understanding the Value of an LLM-Based Voting Advice Application Chatbot

Voting advice applications (VAAs), which have become increasingly prominent in European elections, are seen as a successful tool for boosting electorates' political knowledge and engagement. However, VAAs' complex language and rigid presentation constrain their utility to less-sophisticated voters. While previous work enhanced VAAs' click-based interaction with scripted explanations, a conversational chatbot's potential for tailored discussion and deliberate political decision-making remains untapped. Our exploratory mixed-method study investigates how LLM-based chatbots can support voting preparation. We deployed a VAA chatbot to 331 users before Germany's 2024 European Parliament election, gathering insights from surveys, conversation logs, and 10 follow-up interviews. Participants found the VAA chatbot intuitive and informative, citing its simple language and flexible interaction. We further uncovered VAA chatbots' role as a catalyst for reflection and rationalization. Expanding on participants' desire for transparency, we provide design recommendations for building interactive and trustworthy VAA chatbots.

cs.HC↗

arXiv2025

A Weak Supervision Learning Approach Towards an Equitable Mobility Estimation

The scarcity and high cost of labeled high-resolution imagery have long challenged remote sensing applications, particularly in low-income regions where high-resolution data are scarce. In this study, we propose a weak supervision framework that estimates parking lot occupancy using 3m resolution satellite imagery. By leveraging coarse temporal labels -- based on the assumption that parking lots of major supermarkets and hardware stores in Germany are typically full on Saturdays and empty on Sundays -- we train a pairwise comparison model that achieves an AUC of 0.92 on large parking lots. The proposed approach minimizes the reliance on expensive high-resolution images and holds promise for scalable urban mobility analysis. Moreover, the method can be adapted to assess transit patterns and resource allocation in vulnerable communities, providing a data-driven basis to improve the well-being of those most in need.

cs.CV↗

arXiv2025

Coverage Biases in High-Resolution Satellite Imagery

Satellite imagery is increasingly used to complement traditional data collection approaches such as surveys and censuses across scientific disciplines. However, we ask: Do all places on earth benefit equally from this new wealth of information? In this study, we investigate coverage bias of major satellite constellations that provide optical satellite imagery with a ground sampling distance below 10 meters, evaluating both the future on-demand tasking opportunities as well as the availability of historic images across the globe. Specifically, forward-looking, we estimate how often different places are revisited during a window of 30 days based on the satellites' orbital paths, thus investigating potential coverage biases caused by physical factors. We find that locations farther away from the equator are generally revisited more frequently by the constellations under study. Backward-looking, we show that historic satellite image availability -- based on metadata collected from major satellite imagery providers -- is influenced by socio-economic factors on the ground: less developed, less populated places have less satellite images available. Furthermore, in three small case studies on recent conflict regions in this world, namely Gaza, Sudan and Ukraine, we show that also geopolitical events play an important role in satellite image availability, hinting at underlying business model decisions. These insights lay bare that the digital dividend yielded by satellite imagery is not equally distributed across our planet.

cs.CY↗