Major depression constitutes a serious challenge in personal and public health. Tens of millions people each year suffer from only fraction receives adequate treatment. We explore the potential to use social media detect diagnose major depressive disorder individuals. first employ crowdsourcing compile set Twitter users who report being diagnosed with clinical depression, based on standard psychometric instrument. Through their postings over preceding onset we measure behavioral attributes...

Recent debate has centered on the relative promise of focusing user-interface research developing new metaphors and tools that enhance users abilities to directly manipulate objects versus directing effort toward interface agents provide automation. In this paper, we review principles show for allowing engineers human-computer interaction through an elegant coupling automated services with direct manipulation. Key ideas will be highlighted in terms Lookout system scheduling meeting management.

Advances in artificial intelligence (AI) frame opportunities and challenges for user interface design. Principles human-AI interaction have been discussed the human-computer community over two decades, but more study innovation are needed light of advances AI growing uses technologies human-facing applications. We propose 18 generally applicable design guidelines interaction. These validated through multiple rounds evaluation including a with 49 practitioners who tested against 20 popular...

Artificial intelligence (AI) researchers have been developing and refining large language models (LLMs) that exhibit remarkable capabilities across a variety of domains tasks, challenging our understanding learning cognition. The latest model developed by OpenAI, GPT-4, was trained using an unprecedented scale compute data. In this paper, we report on investigation early version when it still in active development OpenAI. We contend (this of) GPT-4 is part new cohort LLMs (along with ChatGPT...

We formulate and study search algorithms that consider a user's prior interactions with wide variety of content to personalize current Web search. Rather than relying on the unrealistic assumption people will precisely specify their intent when searching, we pursue techniques leverage implicit information about interests. This is used re-rank results within relevance feedback framework. explore rich models user interests, built from both search-related information, such as previously issued...

We report on a diary study of the activities information workers aimed at characterizing how people interleave multiple tasks amidst interruptions. The week-long revealed type and complexity performed, nature interruptions experienced, difficulty shifting among numerous tasks. present key findings from discuss implications findings. Finally, we describe promising directions in design software tools for task management, motivated by

We present a study of anonymized data capturing month high-level communication activities within the whole Microsoft Messenger instant-messaging system. examine characteristics and patterns that emerge from collective dynamics large numbers people, rather than actions individuals. The dataset contains summary properties 30 billion conversations among 240 million people. From data, we construct graph with 180 nodes 1.3 undirected edges, creating largest social network constructed analyzed to...

Article Free Access Share on Sensing techniques for mobile interaction Authors: Ken Hinckley Microsoft Research, One Way, Redmond, WA WAView Profile , Jeff Pierce Mike Sinclair Eric Horvitz Authors Info & Claims UIST '00: Proceedings of the 13th annual ACM symposium User interface software and technologyNovember 2000 Pages 91–100https://doi.org/10.1145/354401.354417Online:01 November 2000Publication History 336citation6,833DownloadsMetricsTotal Citations336Total Downloads6,833Last 12...

Physical activity helps people maintain a healthy weight and reduces the risk for several chronic diseases. Although this knowledge is widely recognized, adults children in many countries around world do not get recommended amounts of physical activity. interventions are found to be ineffective at increasing or reaching inactive populations, there have been anecdotal reports increased due novel mobile games that embed game play world. The most recent salient example such Pokémon Go, which...

Depression is a serious and widespread public health challenge. We examine the potential for leveraging social media postings as new type of lens in understanding depression populations. Information gleaned from bears to complement traditional survey techniques its ability provide finer grained measurements over time while radically expanding population sample sizes. present work on using crowdsourcing methodology build large corpus Twitter that have been shared by individuals diagnosed with...

The World Wide Web provides an abundant source of medical information. This information can assist people who are not healthcare professionals to better understand health and illness, provide them with feasible explanations for symptoms. However, the has potential increase anxieties have little or no training, especially when search is employed as a diagnostic procedure. We use term cyberchondria refer unfounded escalation concerns about common symptomatology, based on review results...

We consider social media as a promising tool for public health, focusing on the use of Twitter posts to build predictive models about forthcoming influence childbirth behavior and mood new mothers. Using posts, we quantify postpartum changes in 376 mothers along dimensions engagement, emotion, network, linguistic style. then construct statistical from training set observations these measures before after reported childbirth, forecast significant The can classify who will change significantly...

Studies of search habits reveal that people engage in many tasks involving collaboration with others, such as travel planning, organizing social events, or working on a homework assignment. However, current Web tools are designed for single user, alone. We introduce SearchTogether, prototype enables groups remote users to synchronously asynchronously collaborate when searching the Web. describe an example usage scenario, and discuss ways SearchTogether facilitates by supporting awareness,...

We explore the challenge of preserving patients' privacy in electronic health record systems. argue that security such systems should be enforced via encryption as well access control. Furthermore, we for approaches enable patients to generate and store keys, so is protected host data center compromised. The standard argument against an approach would interfere with functionality system. However, show can build efficient system allows both share partial rights others, perform searches over...

Abstract: Pathfinder is an expert system that assists surgical pathologists with the diagnosis of lymph-node diseases. The program one a growing number normative systems use probability and decision theory to acquire, represent, manipulate, explain uncertain medical knowledge. In this article, we describe our research in uncertain-reasoning paradigms was stimulated by development program. We discuss limitations early decision-theoretic methods for reasoning under uncertainty initial attempts...

Large language models (LLMs) have demonstrated remarkable capabilities in natural understanding and generation across various domains, including medicine. We present a comprehensive evaluation of GPT-4, state-of-the-art LLM, on medical competency examinations benchmark datasets. GPT-4 is general-purpose model that not specialized for problems through training or engineered to solve clinical tasks. Our analysis covers two sets official practice materials the USMLE, three-step examination...

The birth of a child is major milestone in the life parents. We leverage Facebook data shared voluntarily by 165 new mothers as streams evidence for characterizing their postnatal experiences. consider multiple measures including activity, social capital, emotion, and linguistic style participants' pre- periods. Our study includes detecting predicting onset post-partum depression (PPD). work complements recent on significant postpartum changes behavior, language, affect from Twitter data. In...

We show how machine learning and inference can be harnessed to leverage the complementary strengths of humans computational agents solve crowdsourcing tasks. construct a set Bayesian predictive models from data describe operate within an overall crowd-sourcing architecture that combines efforts people vision on task classifying celestial bodies defined citizens' science project named Galaxy Zoo. learned probabilistic used fuse human contributions predict behaviors workers. employ multiple...

Decisions made by human-AI teams (e.g., AI-advised humans) are increasingly common in high-stakes domains such as healthcare, criminal justice, and finance. Achieving high team performance depends on more than just the accuracy of AI system: Since human may have different expertise, highest is often reached when they both know how to complement one another. We focus a factor that crucial supporting complementary: human’s mental model capabilities, specifically system’s error boundary (i.e....