Siddharth Suri

Chirag Shah, Ryen W. White, Reid Andersen et al.

Log data can reveal valuable information about how users interact with Web search services, what they want, and how satisfied they are. However, analyzing user intents in log data is not easy, especially for emerging forms of Web search such as AI-driven chat. To understand user intents from log data, we need a way to label them with meaningful categories that capture their diversity and dynamics. Existing methods rely on manual or machine-learned labeling, which are either expensive or inflexible for large and dynamic datasets. We propose a novel solution using large language models (LLMs), which can generate rich and relevant concepts, descriptions, and examples for user intents. However, using LLMs to generate a user intent taxonomy and apply it for log analysis can be problematic for two main reasons: (1) such a taxonomy is not externally validated; and (2) there may be an undesirable feedback loop. To address this, we propose a new methodology with human experts and assessors to verify the quality of the LLM-generated taxonomy. We also present an end-to-end pipeline that uses an LLM with human-in-the-loop to produce, refine, and apply labels for user intent analysis in log data. We demonstrate its effectiveness by uncovering new insights into user intents from search and chat logs from the Microsoft Bing commercial search engine. The proposed work's novelty stems from the method for generating purpose-driven user intent taxonomies with strong validation. This method not only helps remove methodological and practical bottlenecks from intent-focused research, but also provides a new framework for generating, validating, and applying other kinds of taxonomies in a scalable and adaptable way with reasonable human effort.

Yulin Yu, Yizhou Li, Siddharth Suri et al.

Conversational generative AI systems such as ChatGPT are transforming how people seek and engage with information online. Unlike traditional search engines, these systems support open-ended, conversational inquiry, yet it remains unclear whether they ultimately expand or constrain the diversity of knowledge that users encounter in online search spaces, a primary foundation for knowledge work, learning, and innovation. Using over 200,000 real-world human-ChatGPT interactions, we examine how generative-AI-mediated inquiry reshapes diversity in both user inputs and system outputs through the lens of searchability - whether queries could plausibly be answered by traditional search engines. We find that almost 80% of ChatGPT user queries are non-searchable and span a broader knowledge space and topics than searchable queries, indicating expanded modes of inquiry. However, for comparable searchable queries, AI responses are less diverse than Google search results in the majority of topics. Moreover, the diversity of AI responses predicts subsequent changes in users' inquiry diversity, revealing a feedback loop between AI outputs and human exploration. These findings highlight a tension between expanded inquiry and constrained information exposure, with implications for designing hybrid search and generative-AI systems that better support exploratory knowledge seeking.

Scott Counts, Yan Chen, Jing Dong et al.

M365 Copilot is used every week by millions of people across more than a million companies around the world as part of their workflows. Uniquely positioned in the AI landscape given its near-exclusive use for work purposes, M365 Copilot can offer a clear picture of how people use AI for work and where that usage may expand next. This paper characterizes that usage through direct classification of user interactions with M365 Copilot Chat. Based on an anonymized and privacy-preserving analysis of a sample of approximately 5.5 million sessions, we combine a learned classification of user intent with a classification of O*NET work activities done with M365 Copilot Chat. We find that M365 Copilot is emerging as an everyday assistant for knowledge work: writing dominates, but users also rely on it for information retrieval, analysis, decision making and strategizing, and evaluating and diagnosing programs and systems, among others. Information seeking tasks remain common, but time trends suggest a relative shift away from ``chat as search'' and toward content and communication-related work. Comparisons across occupational groupings and to work done in the labor market further show that usage is broad but uneven, where the relative share of work done with M365 Copilot Chat cuts across jobs in some cases and is occupation-specific in others. Areas of relative underrepresentation in the labor market suggest the next frontier for enterprise AI adoption.

Mengting Wan, Tara Safavi, Sujay Kumar Jauhar et al.

Transforming unstructured text into structured and meaningful forms, organized by useful category labels, is a fundamental step in text mining for downstream analysis and application. However, most existing methods for producing label taxonomies and building text-based label classifiers still rely heavily on domain expertise and manual curation, making the process expensive and time-consuming. This is particularly challenging when the label space is under-specified and large-scale data annotations are unavailable. In this paper, we address these challenges with Large Language Models (LLMs), whose prompt-based interface facilitates the induction and use of large-scale pseudo labels. We propose TnT-LLM, a two-phase framework that employs LLMs to automate the process of end-to-end label generation and assignment with minimal human effort for any given use-case. In the first phase, we introduce a zero-shot, multi-stage reasoning approach which enables LLMs to produce and refine a label taxonomy iteratively. In the second phase, LLMs are used as data labelers that yield training samples so that lightweight supervised classifiers can be reliably built, deployed, and served at scale. We apply TnT-LLM to the analysis of user intent and conversational domain for Bing Copilot (formerly Bing Chat), an open-domain chat-based search engine. Extensive experiments using both human and automatic evaluation metrics demonstrate that TnT-LLM generates more accurate and relevant label taxonomies when compared against state-of-the-art baselines, and achieves a favorable balance between accuracy and efficiency for classification at scale. We also share our practical experiences and insights on the challenges and opportunities of using LLMs for large-scale text mining in real-world applications.

Kiran Tomlinson, Sonia Jaffe, Will Wang et al.

Given the rapid adoption of generative AI and its potential to impact a wide range of tasks, understanding the effects of AI on the economy is one of society's most important questions. In this work, we take a step toward that goal by analyzing the work activities people do with AI, how successfully and broadly those activities are done, and combine that with data on what occupations do those activities. We analyze a dataset of 200k anonymized and privacy-scrubbed conversations between users and Microsoft Bing Copilot, a publicly available generative AI system. We find the most common work activities people seek AI assistance for involve gathering information and writing, while the most common activities that AI itself is performing are providing information and assistance, writing, teaching, and advising. Combining these activity classifications with measurements of task success and scope of impact, we compute an AI applicability score for each occupation. We find the highest AI applicability scores for knowledge work occupation groups such as computer and mathematical, and office and administrative support, as well as occupations such as sales whose work activities involve providing and communicating information. Additionally, we characterize the types of work activities performed most successfully, how wage and education correlate with AI applicability, and how real-world usage compares to predictions of occupational AI impact.

Siddharth Suri, Scott Counts, Leijie Wang et al.

Until recently, search engines were the predominant method for people to access online information. The recent emergence of large language models (LLMs) has given machines new capabilities such as the ability to generate new digital artifacts like text, images, code etc., resulting in a new tool, a generative search engine, which combines the capabilities of LLMs with a traditional search engine. Through the empirical analysis of Bing Copilot (Bing Chat), one of the first publicly available generative search engines, we analyze the types and complexity of tasks that people use Bing Copilot for compared to Bing Search. Findings indicate that people use the generative search engine for more knowledge work tasks that are higher in cognitive complexity than were commonly done with a traditional search engine.

Ying-Chun Lin, Jennifer Neville, Jack W. Stokes et al.

Accurate and interpretable user satisfaction estimation (USE) is critical for understanding, evaluating, and continuously improving conversational systems. Users express their satisfaction or dissatisfaction with diverse conversational patterns in both general-purpose (ChatGPT and Bing Copilot) and task-oriented (customer service chatbot) conversational systems. Existing approaches based on featurized ML models or text embeddings fall short in extracting generalizable patterns and are hard to interpret. In this work, we show that LLMs can extract interpretable signals of user satisfaction from their natural language utterances more effectively than embedding-based approaches. Moreover, an LLM can be tailored for USE via an iterative prompting framework using supervision from labeled examples. The resulting method, Supervised Prompting for User satisfaction Rubrics (SPUR), not only has higher accuracy but is more interpretable as it scores user satisfaction via learned rubrics with a detailed breakdown.

Carlos Toxtli, Siddharth Suri, Saiph Savage

Crowdsourcing markets provide workers with a centralized place to find paid work. What may not be obvious at first glance is that, in addition to the work they do for pay, crowd workers also have to shoulder a variety of unpaid invisible labor in these markets, which ultimately reduces workers' hourly wages. Invisible labor includes finding good tasks, messaging requesters, or managing payments. However, we currently know little about how much time crowd workers actually spend on invisible labor or how much it costs them economically. To ensure a fair and equitable future for crowd work, we need to be certain that workers are being paid fairly for all of the work they do. In this paper, we conduct a field study to quantify the invisible labor in crowd work. We build a plugin to record the amount of time that 100 workers on Amazon Mechanical Turk dedicate to invisible labor while completing 40,903 tasks. If we ignore the time workers spent on invisible labor, workers' median hourly wage was $3.76. But, we estimated that crowd workers in our study spent 33% of their time daily on invisible labor, dropping their median hourly wage to $2.83. We found that the invisible labor differentially impacts workers depending on their skill level and workers' demographics. The invisible labor category that took the most time and that was also the most common revolved around workers having to manage their payments. The second most time-consuming invisible labor category involved hyper-vigilance, where workers vigilantly watched over requesters' profiles for newly posted work or vigilantly searched for labor. We hope that through our paper, the invisible labor in crowdsourcing becomes more visible, and our results help to reveal the larger implications of the continuing invisibility of labor in crowdsourcing.

Longqi Yang, Sonia Jaffe, David Holtz et al.

The COVID-19 pandemic has had a wide-ranging impact on information workers such as higher stress levels, increased workloads, new workstreams, and more caregiving responsibilities during lockdown. COVID-19 also caused the overwhelming majority of information workers to rapidly shift to working from home (WFH). The central question this work addresses is: can we isolate the effects of WFH on information workers' collaboration activities from all other factors, especially the other effects of COVID-19? This is important because in the future, WFH will likely to be more common than it was prior to the pandemic. We use difference-in-differences (DiD), a causal identification strategy commonly used in the social sciences, to control for unobserved confounding factors and estimate the causal effect of WFH. Our analysis relies on measuring the difference in changes between those who WFH prior to COVID-19 and those who did not. Our preliminary results suggest that on average, people spent more time on collaboration in April (Post WFH mandate) than in February (Pre WFH mandate), but this is primarily due to factors other than WFH, such as lockdowns during the pandemic. The change attributable to WFH specifically is in the opposite direction: less time on collaboration and more focus time. This reversal shows the importance of using causal inference: a simple analysis would have resulted in the wrong conclusion. We further find that the effect of WFH is moderated by individual remote collaboration experience prior to WFH. Meanwhile, the medium for collaboration has also shifted due to WFH: instant messages were used more, whereas scheduled meetings were used less. We discuss design implications -- how future WFH may affect focused work, collaborative work, and creative work.

Andi Peng, Besmira Nushi, Emre Kiciman et al.

Although systematic biases in decision-making are widely documented, the ways in which they emerge from different sources is less understood. We present a controlled experimental platform to study gender bias in hiring by decoupling the effect of world distribution (the gender breakdown of candidates in a specific profession) from bias in human decision-making. We explore the effectiveness of \textit{representation criteria}, fixed proportional display of candidates, as an intervention strategy for mitigation of gender bias by conducting experiments measuring human decision-makers' rankings for who they would recommend as potential hires. Experiments across professions with varying gender proportions show that balancing gender representation in candidate slates can correct biases for some professions where the world distribution is skewed, although doing so has no impact on other professions where human persistent preferences are at play. We show that the gender of the decision-maker, complexity of the decision-making task and over- and under-representation of genders in the candidate slate can all impact the final decision. By decoupling sources of bias, we can better isolate strategies for bias mitigation in human-in-the-loop systems.