Recent advancements in large language models (LLMs) have revolutionized natural language processing (NLP) and expanded their applications across diverse domains. However, despite their impressive capabilities, LLMs have been shown to reflect and perpetuate harmful societal biases, including those based on ethnicity, gender, and religion. A critical and underexplored issue is the reinforcement of caste-based biases, particularly towards India's marginalized caste groups such as Dalits and Shudras. In this paper, we address this gap by proposing DECASTE, a novel, multi-dimensional framework designed to detect and assess both implicit and explicit caste biases in LLMs. Our approach evaluates caste fairness across four dimensions: socio-cultural, economic, educational, and political, using a range of customized prompting strategies. By benchmarking several state-of-the-art LLMs, we reveal that these models systematically reinforce caste biases, with significant disparities observed in the treatment of oppressed versus dominant caste groups. For example, bias scores are notably elevated when comparing Dalits and Shudras with dominant caste groups, reflecting societal prejudices that persist in model outputs. These results expose the subtle yet pervasive caste biases in LLMs and emphasize the need for more comprehensive and inclusive bias evaluation methodologies that assess the potential risks of deploying such models in real-world contexts.

Rapid technological innovation threatens to leave much of the global workforce behind. Today's economy juxtaposes white-hot demand for skilled labor against stagnant employment prospects for workers unprepared to participate in a digital economy. It is a moment of peril and opportunity for every country, with outcomes measured in long-term capital allocation and the life satisfaction of billions of workers. To meet the moment, governments and markets must find ways to quicken the rate at which the supply of skills reacts to changes in demand. More fully and quickly understanding labor market intelligence is one route. In this work, we explore the utility of time series forecasts to enhance the value of skill demand data gathered from online job advertisements. This paper presents a pipeline which makes one-shot multi-step forecasts into the future using a decade of monthly skill demand observations based on a set of recurrent neural network methods. We compare the performance of a multivariate model versus a univariate one, analyze how correlation between skills can influence multivariate model results, and present predictions of demand for a selection of skills practiced by workers in the information technology industry.

Learning never stops for successful workers, who must grow their careers while coping with the changing expectations of employers. Robust job-skill representations can empower workers by helping them to better decipher viable job changes given their current skill set and guide them toward skills they can learn to meet career goals. In this work we combine threads of research in economics and AI to improve upon existing job-skill representation methodology and performance. We build a benchmark dataset of between-job transitions from US Census data and show that a representation trained on a large set of online job postings via a transformer-based architecture outperforms existing baselines. Further analysis demonstrates that this model is better able to transfer across taxonomies than existing models.

Designing systems to offer support to experts during cognitive intensive tasks at the right time is still a challenging endeavor, despite years of research progress in the area. This paper proposes a gaze model based on eye tracking empirical data to identify when a system should proactively interact with the expert during visual inspection tasks. The gaze model derives from the analyses of a user study where 11 seismic interpreters were asked to perform the visual inspection task of seismic images from known and unknown basins. The eye tracking fixation patterns were triangulated with pupil dilations and thinking-aloud data. Results show that cumulative saccadic distances allow identifying when additional information could be offered to support seismic interpreters, changing the visual search behavior from exploratory to goal-directed.

Many studies have evaluated different uses of collaborative software. Typically, the research has focused on the shortcomings and, sometimes, the ways end users succeed or fail to work around these shortcomings. In a recent field study, surprisingly, a group demonstrated unimpaired dexterity using a full range of collaborative software. Some interesting lessons emerged from observing these “perfect ” collaborators. Lessons include implications for more typical or “less than perfect ” end users, especially around the adoption of collaboration technology. Also, there is a general, but subtle, lesson that studying successful users of technology (or “ideal end users ” as we put it) can be as valuable as studying those who struggle with technology and highlight its shortcomings.

In our research, we have been concerned with the question of how to make relevant features of security situations visible to users in order to allow them to make informed decisions regarding potential privacy and security problems, as well as regarding potential implications of their actions. To this end, we have designed technical infrastructures that make visible the configurations, activities, and implications of available security mechanisms. This thus allows users to make informed choices and take coordinated and appropriate actions when necessary. This work differs from the more traditional security usability work in that our focus is not only on the usability of security mechanism (e.g., the ease-of-use of an access control interface), but how security can manifest itself as part of people's interactions with and through information systems (i.e., how people experience and interpret privacy and security situations, and are enabled or constrained by existing technological mechanisms to act appropriately). In this paper, we report our experiences designing, developing, and testing two technical infrastructures for supporting this approach for usable security.