A5087306226- Speech and dialogue systems
- Mobile Health and mHealth Applications
- Machine Learning in Healthcare
- Digital Mental Health Interventions
- Natural Language Processing Techniques
- Topic Modeling
- Electronic Health Records Systems
- Chronic Disease Management Strategies
- Mental Health Research Topics
- Advanced Text Analysis Techniques
- Physical Activity and Health
- Nutritional Studies and Diet
- Diabetes Management and Education
- Sentiment Analysis and Opinion Mining
- Biomedical Text Mining and Ontologies
- Artificial Intelligence in Healthcare and Education
- Impact of Technology on Adolescents
- Privacy-Preserving Technologies in Data
- Behavioral Health and Interventions
- Anomaly Detection Techniques and Applications
- Health Policy Implementation Science
- Clinical practice guidelines implementation
- Cloud Data Security Solutions
- LGBTQ Health, Identity, and Policy
- Security in Wireless Sensor Networks
Pfizer (United States)
2023
IBM (United States)
2009-2021
Enzo Life Sciences (United States)
2020
IBM Research - Thomas J. Watson Research Center
2009-2019
Poznań University of Economics and Business
2014
San Jose State University
2014
University of Edinburgh
2006-2009
Annotation acquisition is an essential step in training supervised classifiers. However, manual annotation often time-consuming and expensive. The possibility of recruiting annotators through Internet services (e.g., Amazon Mechanic Turk) appealing option that allows multiple labeling tasks to be outsourced bulk, typically with low overall costs fast completion rates. In this paper, we consider the difficult problem classifying sentiment political blog snippets. data from both expert a...
Summary Objectives: The understanding of how stress influences health behavior can provide insights into developing healthy lifestyle interventions. This is traditionally attained through observational studies that examine associations at a population level. nomothetic approach, however, fundamentally limited by the fact environment- person milieu constitutes exposure and experience vary substantially between individuals, modifiable elements these exposures experiences are...
In this paper, we analyze complex gaze tracking data in a collaborative task and apply machine learning models to automatically predict skill-level differences between participants. Specifically, present findings that address the two primary challenges for prediction task: (1) extracting meaningful features from information, (2) casting as (ML) problem. The results show our approach based on profile hidden Markov are up 96% accurate can make determination fast one minute into collaboration,...
This study concerns how to segment a scenario-driven multiparty dialogue and label these segments automatically. We apply approaches that have been proposed for identifying topic boundaries at coarser level the problem of agenda-based in scenario-based meetings. also develop conditional models classify into classes. Experiments segmentation show supervised classification approach combines lexical conversational features outperforms unsupervised chain-based approach, achieving 20% 12%...
Advances in multimedia technologies have enabled the creation of huge archives audio-video recordings meetings, and there is burgeoning interest developing meeting browsers to help users better leverage these archives. A recent study has shown that extractive summaries provide a more efficient way navigating content than simply reading through transcript using record, or via keyword search (Murray, 2007). The summary technique identifies informative dialogue acts generate general purpose...
Owing to advances in sensor technologies on wearable devices, it is feasible measure physical activity of an individual continuously over a long period. These devices afford opportunities understand behaviors, which may then provide basis for tailored behavior interventions. The large volume data however poses challenges management and analysis. We propose novel quantile coarsening analysis (QCA) daily data, with goal reduce the while preserving key information. applied QCA longitudinal...
AMI Meeting Facilitator is a system that performs topic segmentation and extractive summarisation. It consists of three components: (1) segmenter divides meeting into number locally coherent segments, (2) summarizer selects the most important utterances from transcripts, (3) compression component removes less words each utterance based on degree user specified. The goal two-fold: first, we want to provide sufficient visual aids for users interpret what going in recorded meeting; second,...
Offering personalized services through dynamically formed ecosystems is essential to personal wellness management. In this paper, we present the design of a cloud-enabled platform facilitate collection and delivery evidence for personalization in multi-provider ecosystem environment.
Social media platforms have become popular online environments for patients seeking and sharing treatment experiences. These enable us to move beyond traditional sources of clinical information learning about a patient's long-term adherence treatment. While has been studied using data derived from medical records structured surveys, these approaches are limited in that they often 1) time consuming, 2) scale, or 3) lack self-reported patient In this paper, we investigate through discussion...
In this paper, we present Greenolive, an open platform for wellness management ecosystem. Wellness applications, which facilitate preventive care and chronic disease treatments, are considered as a key component to enhance healthcare quality reduce cost. Currently, most of the applications device-oriented, stand-alone software lack APIs that allow new value-added be developed rapidly. Further, these have not fully utilized collected monitoring data generate knowledge can help people further...
In this paper, we explore a novel way to leverage audio information for unsupervised segmentation of multiparty dialogue. Our system which segments directly on patterns derived from sources is evaluated with previous work that lexical found in transcripts. We examine the effectiveness both systems recovering two-layer structure meeting demonstrate audio-based performs significantly better than word-based task. particular, it effectively recover off-topic discussion. Results are encouraging...
Creation of a personalized adherence feedback loop is crucial for initiating and sustaining health behavior change. However, self reports are not sufficient to measure actual adherence. Recording recognizing personal activities in ubiquitous environment has thus emerged as promising solution. In this work, we present model-driven sensor data assessment mechanism capable identifying high level adherence-related activity patterns from low signals. The proposed intelligent sensing algorithm can...