Skip to main content
:professors
:professors
all professors

Nina Hubig

Assistant Professor of Explainable Artificial Intelligence

title
Asst. Prof.
email
:back to overview

Publications

TitleYear
Explainable AI Systems Must Be Contestable: Here’s How to Make It Happen
2025
Catarina Moreira, Anna Palatkina, Dacia Braca, Dylan M Walsh, Peter J Leihn, Fang Chen,…
arXiv preprint arXiv:2506.01662
:read more
Authors:Catarina Moreira, Anna Palatkina, Dacia Braca, Dylan M Walsh, Peter J Leihn, Fang Chen, Nina C Hubig
Publication date:06/02/2025
Journal:arXiv preprint arXiv:2506.01662
Description:As AI regulations around the world intensify their focus on system safety, contestability has become a mandatory, yet ill-defined, safeguard. In XAI, "contestability" remains an empty promise: no formal definition exists, no algorithm guarantees it, and practitioners lack concrete guidance to satisfy regulatory requirements. Grounded in a systematic literature review, this paper presents the first rigorous formal definition of contestability in explainable AI, directly aligned with stakeholder requirements and regulatory mandates. We introduce a modular framework of by-design and post-hoc mechanisms spanning human-centered interfaces, technical architectures, legal processes, and organizational workflows. To operationalize our framework, we propose the Contestability Assessment Scale, a composite metric built on more than twenty quantitative criteria. Through multiple case studies across diverse application domains, we reveal where state-of-the-art systems fall short and show how our framework drives targeted improvements. By converting contestability from regulatory theory into a practical framework, our work equips practitioners with the tools to embed genuine recourse and accountability into AI systems.
Resources:[PDF] from arxiv.org
:view less
The emergence of large language models as tools in literature reviews: a large language model-assisted systematic review
2025
Dmitry Scherbakov, Nina Hubig, Vinita Jansari, Alexander Bakumenko, Leslie A Lenert
Journal of the American Medical Informatics Association 32 (6), 1071-1086
:read more
Authors:Dmitry Scherbakov, Nina Hubig, Vinita Jansari, Alexander Bakumenko, Leslie A Lenert
Publication date:06/01/2025
Journal:Journal of the American Medical Informatics Association
Volume:32
Issue:6
Pages:1071-1086
Publisher:Oxford University Press
Description:Objectives This study aims to summarize the usage of large language models (LLMs) in the process of creating a scientific review by looking at the methodological papers that describe the use of LLMs in review automation and the review papers that mention they were made with the support of LLMs. Materials and Methods The search was conducted in June 2024 in PubMed, Scopus, Dimensions, and Google Scholar by human reviewers. Screening and extraction process took place in Covidence with the help of LLM add-on based on the OpenAI GPT-4o model. ChatGPT and Scite.ai were used in cleaning the data, generating the code for figures, and drafting the manuscript. Results Of the 3788 articles retrieved, 172 studies were deemed eligible for the final review. ChatGPT and GPT-based LLM emerged as the most dominant architecture for review …
Cited by:19
Resources:[HTML] from oup.com
:view less
New Frontiers of Human-centered Explainable AI (HCXAI): Participatory Civic AI, Benchmarking LLMs, XAI Hallucinations, and Responsible AI Audits
2025
Upol Ehsan, Elizabeth A Watkins, Philipp Wintersberger, Carina Manger, Nina Hubig, Saip…
Proceedings of the Extended Abstracts of the CHI Conference on Human Factors in Computi…
:read more
Authors:Upol Ehsan, Elizabeth A Watkins, Philipp Wintersberger, Carina Manger, Nina Hubig, Saiph Savage, Justin D Weisz, Andreas Riener
Publication date:04/26/2025
Journal:Proceedings of the Extended Abstracts of the CHI Conference on Human Factors in Computing Systems
Pages:1-6
Description:Explainable AI (XAI) is more than just “opening” the black box — who opens it matters just as much, if not more, as the ways of opening it. Human-centered XAI (HCXAI) advocates that algorithmic transparency alone is not sufficient for making AI explainable. In our fifth CHI workshop on Human-Centered XAI (HCXAI), we shift our focus to new, emerging frontiers of explainability: (1) participatory approaches toward explainability in civic AI applications; (2) addressing hallucinations in LLMs using explainability benchmarks; (3) connecting HCXAI research with Responsible AI practices, algorithmic auditing, and public policy; and (4) improving representation of XAI issues from the Global South. We have built a strong community of HCXAI researchers through our workshop series whose work has made important conceptual, methodological, and technical impact on the field. In this installment, we will push the frontiers of …
:view less
New Frontiers of Human-centered Explainable AI (HCXAI): Participatory Civic AI, Benchmarking LLMs and Hallucinations for XAI, and Responsible AI Audits
2025
Upol Ehsan, Elizabeth Watkins, Philipp Wintersberger, Carina Manger, Nina Hubig, Saiph …
ACM CHI Conference on Human Factors in Computing Systems
:read more
Authors:Upol Ehsan, Elizabeth Watkins, Philipp Wintersberger, Carina Manger, Nina Hubig, Saiph Savage, Justin Weisz, Andreas Riener
Publication date:04/26/2025
Journal:ACM CHI Conference on Human Factors in Computing Systems
Description:Explainable AI (XAI) is more than just" opening" the black box---who opens it matters just as much, if not more, as the ways of opening it. Human-centered XAI (HCXAI) advocates that algorithmic transparency alone is not sufficient for making AI explainable. In our fifth CHI workshop on Human-Centered XAI (HCXAI), we shift our focus to new, emerging frontiers of explainability:(1) participatory approaches toward explainability in civic AI applications;(2) addressing hallucinations in LLMs using explainability benchmarks;(3) connecting HCXAI research with Responsible AI practices, algorithmic auditing, and public policy; and (4) improving representation of XAI issues from the Global South. We have built a strong community of HCXAI researchers through our workshop series whose work has made important conceptual, methodological, and technical impact on the field. In this installment, we will push the frontiers of …
:view less
Active learning pipeline to automatically identify candidate terms for a CDSS ontology—measures, experiments, and performance
2025
Shailesh Alluri, Keerthana Komatineni, Rohan Goli, Nina Hubig, Hua Min, Yang Gong, Dean…
medRxiv, 2025.04. 15.25325868
:read more
Authors:Shailesh Alluri, Keerthana Komatineni, Rohan Goli, Nina Hubig, Hua Min, Yang Gong, Dean F Sittig, David Robinson, Paul Biondich, Adam Wright, Christian Nøhr, Timothy Law, Arild Faxvaag, Richard D Boyce, Ronald Gimbel, Lior Rennert, Xia Jing
Publication date:04/17/2025
Journal:medRxiv
Pages:2025.04. 15.25325868
Publisher:Cold Spring Harbor Laboratory Press
Description:In this paper, we present an active learning framework designed to enhance the automatic identification of key phrases relevant to a Clinical Decision Support System (CDSS) ontology. CDSS ontology can play a critical role in standardizing medical vocabulary and enabling seamless data integration across healthcare systems. Traditional methods for ontology development are manual, labor-intensive, and require significant domain expertise. Our approach combines a Named Entity Recognition (NER) component based on a BiLSTM-CRF model with an active learning loop to progressively incorporate human experts’ feedback to improve its accuracy. We implement uncertainty sampling as our core data selection strategy, prioritizing instances where the model exhibits low confidence for human review. We introduce new uncertainty aggregation methods—KPSum, KPAvg, DOCSum, and DOCAvg—which, in combination with uncertainty measures such as Maximum Token Probability (MTP), Token Entropy (TE), and Margin, are used to calculate document-level confidence scores. These methods improve the selection process in Active Learning, ensuring that the most informative documents are prioritized for annotation. This study underscores the value of active learning in facilitating ontology development.
Resources:[PDF] from medrxiv.org
:view less
Natural Language Processing and Social Determinants of Health in Mental Health Research: AI-Assisted Scoping Review
2025
Dmitry A Scherbakov, Nina C Hubig, Leslie A Lenert, Alexander V Alekseyenko, Jihad S Obeid
JMIR Mental Health 12 (1), e67192
:read more
Authors:Dmitry A Scherbakov, Nina C Hubig, Leslie A Lenert, Alexander V Alekseyenko, Jihad S Obeid
Publication date:01/16/2025
Journal:JMIR Mental Health
Volume:12
Issue:1
Pages:e67192
Publisher:JMIR Publications Inc., Toronto, Canada
Description:Background: The use of natural language processing (NLP) in mental health research is increasing, with a wide range of applications and datasets being investigated. Objective: This review aims to summarize the use of NLP in mental health research, with a special focus on the types of text datasets and the use of social determinants of health (SDOH) in NLP projects related to mental health. Methods: The search was conducted in September 2024 using a broad search strategy in PubMed, Scopus, and CINAHL Complete. All citations were uploaded to Covidence (Veritas Health Innovation) software. The screening and extraction process took place in Covidence with the help of a custom large language model (LLM) module developed by our team. This LLM module was calibrated and tuned to automate many aspects of the review process. Results: The screening process, assisted by the custom LLM, led to the inclusion of 1768 studies in the final review. Most of the reviewed studies (n= 665, 42.8%) used clinical data as their primary text dataset, followed by social media datasets (n= 523, 33.7%). The United States contributed the highest number of studies (n= 568, 36.6%), with depression (n= 438, 28.2%) and suicide (n= 240, 15.5%) being the most frequently investigated mental health issues. Traditional demographic variables, such as age (n= 877, 56.5%) and gender (n= 760, 49%), were commonly extracted, while SDOH factors were less frequently reported, with urban or rural status being the most used (n= 19, 1.2%). Over half of the citations (n= 826, 53.2%) did not provide clear information on dataset accessibility, although a sizable number …
Cited by:3
Resources:[HTML] from jmir.org
:view less
Improving the adaptive and continuous learning capabilities of artificial neural networks: Lessons from multi-neuromodulatory dynamics
2025
Jie Mei, Alejandro Rodriguez-Garcia, Daigo Takeuchi, Gabriel Wainstein, Nina Hubig, Yal…
arXiv preprint arXiv:2501.06762
:read more
Authors:Jie Mei, Alejandro Rodriguez-Garcia, Daigo Takeuchi, Gabriel Wainstein, Nina Hubig, Yalda Mohsenzadeh, Srikanth Ramaswamy
Publication date:01/12/2025
Journal:arXiv preprint arXiv:2501.06762
Description:Continuous, adaptive learning-the ability to adapt to the environment and improve performance-is a hallmark of both natural and artificial intelligence. Biological organisms excel in acquiring, transferring, and retaining knowledge while adapting to dynamic environments, making them a rich source of inspiration for artificial neural networks (ANNs). This study explores how neuromodulation, a fundamental feature of biological learning systems, can help address challenges such as catastrophic forgetting and enhance the robustness of ANNs in continuous learning scenarios. Driven by neuromodulators including dopamine (DA), acetylcholine (ACh), serotonin (5-HT) and noradrenaline (NA), neuromodulatory processes in the brain operate at multiple scales, facilitating dynamic responses to environmental changes through mechanisms ranging from local synaptic plasticity to global network-wide adaptability. Importantly, the relationship between neuromodulators, and their interplay in the modulation of sensory and cognitive processes are more complex than expected, demonstrating a "many-to-one" neuromodulator-to-task mapping. To inspire the design of novel neuromodulation-aware learning rules, we highlight (i) how multi-neuromodulatory interactions enrich single-neuromodulator-driven learning, (ii) the impact of neuromodulators at multiple spatial and temporal scales, and correspondingly, (iii) strategies to integrate neuromodulated learning into or approximate it in ANNs. To illustrate these principles, we present a case study to demonstrate how neuromodulation-inspired mechanisms, such as DA-driven reward processing and NA-based …
Cited by:2
Resources:[PDF] from arxiv.org
:view less
ETHICS AND ACCEPTANCE OF AI
2024
Shelia Cotten, Aishah Khan, Catherine Mobley, Xia Jing, Jihad Obeid, Grant Goodrich, Ha…
Innovation in Aging 8 (Suppl 1), 280
:read more
Authors:Shelia Cotten, Aishah Khan, Catherine Mobley, Xia Jing, Jihad Obeid, Grant Goodrich, Hamilton Baker, Nina Hubig
Publication date:12/31/2024
Journal:Innovation in Aging
Volume:8
Issue:Suppl 1
Pages:280
Description:Over the past 18 months, artificial intelligence (AI) applications have proliferated and been disseminated across wider audiences than in prior time periods. AI tools are increasingly being used as part of new biomedical devices and may become more integrated into healthcare processes over time. While AI applications may have significant value in healthcare and helping support older adults and individuals across the life course through a variety of mechanisms, if individuals do not perceive these tools as ethical and are not accepting of these new tools, they are likely to fail. Cotten and colleagues report the results of a scoping review designed to assess stakeholder (e.g., patients and healthcare providers) acceptance and willingness to use AI-enabled devices in diagnosis, treatment, and rehabilitation settings. The results of this study identify perceived challenges of integrating AI into healthcare across …
:view less
A Robust Adversarial Ensemble with Causal (Feature Interaction) Interpretations for Image Classification
2024
Chunheng Zhao, Pierluigi Pisu, Gurcan Comert, Negash Begashaw, Varghese Vaidyan, Nina C…
arXiv preprint arXiv:2412.20025
:read more
Authors:Chunheng Zhao, Pierluigi Pisu, Gurcan Comert, Negash Begashaw, Varghese Vaidyan, Nina Christine Hubig
Publication date:12/28/2024
Journal:arXiv preprint arXiv:2412.20025
Description:Deep learning-based discriminative classifiers, despite their remarkable success, remain vulnerable to adversarial examples that can mislead model predictions. While adversarial training can enhance robustness, it fails to address the intrinsic vulnerability stemming from the opaque nature of these black-box models. We present a deep ensemble model that combines discriminative features with generative models to achieve both high accuracy and adversarial robustness. Our approach integrates a bottom-level pre-trained discriminative network for feature extraction with a top-level generative classification network that models adversarial input distributions through a deep latent variable model. Using variational Bayes, our model achieves superior robustness against white-box adversarial attacks without adversarial training. Extensive experiments on CIFAR-10 and CIFAR-100 demonstrate our model's superior adversarial robustness. Through evaluations using counterfactual metrics and feature interaction-based metrics, we establish correlations between model interpretability and adversarial robustness. Additionally, preliminary results on Tiny-ImageNet validate our approach's scalability to more complex datasets, offering a practical solution for developing robust image classification models.
Resources:[PDF] from arxiv.org
:view less
Keyphrase Identification Using Minimal Labeled Data with Hierarchical Contexts and Transfer Learning
2024
Rohan Goli, Keerthana Komatineni, Shailesh Alluri, Nina Hubig, Hua Min, Yang Gong, Dean…
medRxiv, 2023.01. 26.23285060
:read more
Authors:Rohan Goli, Keerthana Komatineni, Shailesh Alluri, Nina Hubig, Hua Min, Yang Gong, Dean F Sittig, Lior Rennert, David Robinson, Paul Biondich, Adam Wright, Christian Nøhr, Timothy Law, Arild Faxvaag, Aneesa Weaver, Ronald Gimbel, Xia Jing
Publication date:11/18/2024
Journal:medRxiv
Pages:2023.01. 26.23285060
Description:BackgroundMethods
Cited by:3
Resources:[HTML] from nih.gov
:view less
PhenoMetaboDiff: R Package for Analysis and Visualization of Phenotype Microarray Data
2024
Rini Pauly, Mehtab Iqbal, Narae Lee, Bridgette Allen Moffitt, Sara Moir Sarasua, Luyi L…
Genes 15 (11), 1362
:read more
Authors:Rini Pauly, Mehtab Iqbal, Narae Lee, Bridgette Allen Moffitt, Sara Moir Sarasua, Luyi Li, Nina Christine Hubig, Luigi Boccuto
Publication date:10/24/2024
Journal:Genes
Volume:15
Issue:11
Pages:1362
Publisher:MDPI
Description:Background PhenoMetaboDiff is a novel R package for computational analysis and visualization of data generated by Biolog Phenotype Mammalian Microarrays (PM-Ms). These arrays measure the energy production of mammalian cells in different metabolic environments, assess the metabolic activity of cells exposed to various drugs or energy sources, and compare the metabolic profiles of cells from individuals affected by specific disorders versus healthy controls. Methods PhenoMetaboDiff has several modules that facilitate statistical analysis by sample comparisons using non-parametric Mann–Whitney U-test, the integration of the OPM package (an R package for analysing OmniLog® phenotype microarray data) for robust file conversion, and calculation of slope and area under the curve (AUC). In addition, the built-in visualization allows specific wells to be visualized in selected pathways for a particular time slice. Results Compared to the standard OPM package, the features developed in PhenoMetaboDiff assess metabolic profiles by employing statistical tests and visualize the dynamic nature of the energy production in several conditions. Examples of how this package can be used are demonstrated for several rare disease conditions. The incorporation of a graphical user interface expands the utility of this program to both expert and novice users of R. Conclusions PhenoMetaboDiff makes the deployment of the cutting-edge Biolog system available to any researcher. BackgroundMethodsResultsConclusions
Resources:[HTML] from mdpi.com
:view less
Reusable Generic Clinical Decision Support System Module for Immunization Recommendations in Resource-Constraint Settings
2024
Samuil Orlioglu, Akash Shanmugan Boobalan, Kojo Abanyie, Richard D Boyce, Hua Min, Yang…
medRxiv
:read more
Authors:Samuil Orlioglu, Akash Shanmugan Boobalan, Kojo Abanyie, Richard D Boyce, Hua Min, Yang Gong, Dean F Sittig, Paul Biondich, Adam Wright, Christian Nøhr, Timothy Law, David Robinson, Arild Faxvaag, Nina Hubig, Ronald Gimbel, Lior Rennert, Xia Jing
Publication date:09/23/2024
Journal:medRxiv
Description:Clinical decision support systems (CDSS) are routinely employed in clinical settings to improve quality of care, ensure patient safety, and deliver consistent medical care. However, rule-based CDSS, currently available, do not feature reusable rules. In this study, we present CDSS with reusable rules. Our solution includes a common CDSS module, electronic medical record (EMR) specific adapters, CDSS rules written in the clinical quality language (CQL) (derived from CDC immunization recommendations), and patient records in fast healthcare interoperability resources (FHIR) format. The proposed CDSS is entirely browser-based and reachable within the user’s EMR interface at the client-side. This helps to avoid the transmission of patient data and privacy breaches. Additionally, we propose to provide means of managing and maintaining CDSS rules to allow the end users to modify them independently …
Resources:[HTML] from nih.gov
:view less
Low Resource ICD Coding of Hospital Discharge Summaries
2024
Ashton Williamson, David de Hilster, Amnon Meyers, Nina Hubig, Amy Apon
Proceedings of the 23rd Workshop on Biomedical Natural Language Processing, 548-558
:read more
Authors:Ashton Williamson, David de Hilster, Amnon Meyers, Nina Hubig, Amy Apon
Publication date:08/01/2024
Journal:Proceedings of the 23rd Workshop on Biomedical Natural Language Processing
Pages:548-558
Description:Medical coding is the process by which standardized medical codes are assigned to patient health records. This is a complex and challenging task that typically requires an expert human coder to review health records and assign codes from a classification system based on a standard set of rules. Since health records typically consist of a large proportion of free-text documents, this problem has traditionally been approached as a natural language processing (NLP) task. While machine learning-based methods have seen recent popularity on this task, they tend to struggle with codes that are assigned less frequently, for which little or no training data exists. In this work we utilize the open-source NLP programming language, NLP++, to design and build an automated system to assign International Classification of Diseases (ICD) codes to discharge summaries that functions in the absence of labeled training data. We evaluate our system using the MIMIC-III dataset and find that for codes with little training data, our approach achieves competitive performance compared to state-of-the-art machine learning approaches.
Cited by:1
Resources:[PDF] from aclanthology.org
:view less
Advancing anomaly detection: Non-semantic financial data encoding with llms
2024
Alexander Bakumenko, Kateřina Hlaváčková-Schindler, Claudia Plant, Nina C Hubig
arXiv preprint arXiv:2406.03614
:read more
Authors:Alexander Bakumenko, Kateřina Hlaváčková-Schindler, Claudia Plant, Nina C Hubig
Publication date:06/05/2024
Journal:arXiv preprint arXiv:2406.03614
Description:Detecting anomalies in general ledger data is of utmost importance to ensure trustworthiness of financial records. Financial audits increasingly rely on machine learning (ML) algorithms to identify irregular or potentially fraudulent journal entries, each characterized by a varying number of transactions. In machine learning, heterogeneity in feature dimensions adds significant complexity to data analysis. In this paper, we introduce a novel approach to anomaly detection in financial data using Large Language Models (LLMs) embeddings. To encode non-semantic categorical data from real-world financial records, we tested 3 pre-trained general purpose sentence-transformer models. For the downstream classification task, we implemented and evaluated 5 optimized ML models including Logistic Regression, Random Forest, Gradient Boosting Machines, Support Vector Machines, and Neural Networks. Our experiments demonstrate that LLMs contribute valuable information to anomaly detection as our models outperform the baselines, in selected settings even by a large margin. The findings further underscore the effectiveness of LLMs in enhancing anomaly detection in financial journal entries, particularly by tackling feature sparsity. We discuss a promising perspective on using LLM embeddings for non-semantic data in the financial context and beyond.
Cited by:11
Resources:[PDF] from arxiv.org
:view less
Literature Analysis on Ontologies Applied in Clinical Decision Support Systems
2024
JING Xia, MIN Hua, GONG Yang, Paul Biondich, David Robinson, D LAW Timothy, G NOHR Chri…
Studies in health technology and informatics 310, 1347
:read more
Authors:JING Xia, MIN Hua, GONG Yang, Paul Biondich, David Robinson, D LAW Timothy, G NOHR Christian, Arild Faxvaag, Lior Rennert, Nina C Hubig, Ronald W Gimbel
Publication date:01/25/2024
Journal:Studies in health technology and informatics
Volume:310
Pages:1347
Description:2. Methods We used PubMed, the ACM Digital Library, and the Nursing & Allied Health Database for the literature search. The search strategies include CDSS as a MeSH term and Ontolog* or Rule* appear in title or abstract. Each publication was screened independently by at least 2 biomedical informaticians. 3. Results The literature search generated 1235 distinct publications and we included 81 publications [5]. PubMed covered 90% of these publications. We noticed core sources for the 81 publications. Studies in Health Technology and Informatics, as a critical international venue, contributed 21 of the 81 publications. We also noticed the global locations of contributors and their collaborations. Among 81 publications, 58 (72%) had collaborations, of which 20 had international collaborations. 58 (71.6%) of the papers were funded by grants, 6 of which (10.3%) received grants from more than one country.
Resources:[HTML] from nih.gov
:view less
Active Learning Pipeline to Identify Candidate Terms for a CDSS Ontology
2024
Xia Jing, Rohan Goli, Keerthana Komatineni, Shailesh Alluri, Nina Hubig, Hua Min, Yang …
Digital Health and Informatics Innovations for Sustainable Health Care Systems, 1338-1342
:read more
Authors:Xia Jing, Rohan Goli, Keerthana Komatineni, Shailesh Alluri, Nina Hubig, Hua Min, Yang Gong, Dean F Sittig, Paul Biondich, David Robinson, Christian Nøhr, Arild Faxvaag, Adam Wright, Timothy Law, Lior Rennert, Ronald Gimbel
Publication date:01/01/2024
Journal:Digital Health and Informatics Innovations for Sustainable Health Care Systems
Pages:1338-1342
Publisher:IOS Press
Description:Ontology is essential for achieving health information and information technology application interoperability in the biomedical fields and beyond. Traditionally, ontology construction is carried out manually by human domain experts (HDE). Here, we explore an active learning approach to automatically identify candidate terms from publications, with manual verification later as a part of a deep learning model training and learning process. We introduce the overall architecture of the active learning pipeline and present some preliminary results. This work is a critical and complementary component in addition to manually building the ontology, especially during the long-term maintenance stage.
Cited by:1
Resources:[PDF] from iospress.nl
:view less
Low-resource ICD Coding of Discharge Summaries
2024
Ashton Williamson
:read more
Authors:Ashton Williamson
Publication date:01/01/2024
Description:Medical coding is the process by which standardized medical codes are assigned to patient health records. This is a complex and challenging task that typically requires an expert human coder to review health records and assign codes from a classification system based on a standard set of rules. Considering the downstream use of these codes in statistical analysis, billing, and patient care, improving the accuracy and efficiency of the medical coding process through automation could have a far-reaching impact on the healthcare domain. Since health records typically consist of a large proportion of free-text documents, this problem has traditionally been approached as a natural language processing (NLP) task. While machine learning-based methods have seen recent popularity on this task, they tend to struggle with codes that are assigned less frequently, for which little or no training data exists.
Resources:[PDF] from clemson.edu
:view less
FlowDyn: A daily streamflow prediction pipeline for dynamical deep neural network applications
2023
S Sadeghi Tabas, Nushrat Humaira, S Samadi, Nina C Hubig
Environmental Modelling & Software 170, 105854
:read more
Authors:S Sadeghi Tabas, Nushrat Humaira, S Samadi, Nina C Hubig
Publication date:12/01/2023
Journal:Environmental Modelling & Software
Volume:170
Pages:105854
Publisher:Elsevier
Description:This paper presents a dynamical neural network framework to understand how catchment systems respond to daily rainfall-runoff processes over time. We developed an interactive Python-based deep neural network (DNN) package called FlowDyn (presented through a JS-based web platform) to simulate and forecast daily streamflow data for >180 gauging stations across the globe. Several DNN models, including long short-term memory (LSTM), Gated Recurrent Unit (GRU), and a hybrid network of convolutional neural network and LSTM (ConvLSTM), as well as an auto encoder (AE) model were developed and integrated into the FlowDyn pipeline to analyze and forecast sequential daily streamflow values that are embedded within a web-based application for demonstration and visualization. Inputs were gathered from different web services, including the catchment attributes and meteorology for large-sample …
Cited by:8
:view less
Opening the black box of family-based treatments: An artificial intelligence framework to examine therapeutic alliance and therapist empathy
2023
Phillippe B Cunningham, Jordon Gilmore, Sylvie Naar, Stephanie D Preston, Catherine F E…
Clinical child and family psychology review 26 (4), 975-993
:read more
Authors:Phillippe B Cunningham, Jordon Gilmore, Sylvie Naar, Stephanie D Preston, Catherine F Eubanks, Nina Christina Hubig, Jerome McClendon, Samiran Ghosh, Stacy Ryan-Pettes
Publication date:12/01/2023
Journal:Clinical child and family psychology review
Volume:26
Issue:4
Pages:975-993
Publisher:Springer US
Description:The evidence-based treatment (EBT) movement has primarily focused on core intervention content or treatment fidelity and has largely ignored practitioner skills to manage interpersonal process issues that emerge during treatment, especially with difficult-to-treat adolescents (delinquent, substance-using, medical non-adherence) and those of color. A chief complaint of “real world” practitioners about manualized treatments is the lack of correspondence between following a manual and managing microsocial interpersonal processes (e.g. negative affect) that arise in treating “real world clients.” Although family-based EBTs share core similarities (e.g. focus on family interactions, emphasis on practitioner engagement, family involvement), most of these treatments do not have an evidence base regarding common implementation and treatment process problems that practitioners experience in delivering particular …
Cited by:8
Resources:[PDF] from nih.gov
:view less
DX-FloodLine: End-To-End Deep Explainable Pipeline for Real Time Flood Scene Object Detection From Multimedia Images
2023
Nushrat Humaira, Vidya S Samadi, Nina C Hubig
IEEE Access 11, 110644-110655
:read more
Authors:Nushrat Humaira, Vidya S Samadi, Nina C Hubig
Publication date:10/02/2023
Journal:IEEE Access
Volume:11
Pages:110644-110655
Publisher:IEEE
Description:In recent years, an alarming number of natural disasters, specifically hurricanes and floods have affected the United States. Flooding is a global crisis, however we still do not have an automated approach for real-time flood risk detection and mitigation. Detecting inundation in urban areas and road segments is crucial for vehicle routing and traffic management. Unlike remote sensing images that have been used in past studies, it is better to mine social media data and geo-tagged images or videos for flood estimation in near real time. In this paper, multimedia content was collected from social media streaming services, primarily Twitter for analysis. We propose DX-FloodLine, an interpretable, intelligent, multi-stage, end-to-end, deep neural network-based pipeline to classify near real time emerging flood occurrence from images and detect submerged objects and pedestrians around flooded regions. We introduce a …
Cited by:4
Resources:[PDF] from ieee.org
:view less
Ontologies applied in clinical decision support system rules: Systematic review
2023
Xia Jing, Hua Min, Yang Gong, Paul Biondich, David Robinson, Timothy Law, Christian Noh…
JMIR medical informatics 11, e43053
:read more
Authors:Xia Jing, Hua Min, Yang Gong, Paul Biondich, David Robinson, Timothy Law, Christian Nohr, Arild Faxvaag, Lior Rennert, Nina Hubig, Ronald Gimbel
Publication date:01/19/2023
Journal:JMIR medical informatics
Volume:11
Pages:e43053
Publisher:JMIR Publications
Description:Background Clinical decision support systems (CDSSs) are important for the quality and safety of health care delivery. Although CDSS rules guide CDSS behavior, they are not routinely shared and reused. Objective Ontologies have the potential to promote the reuse of CDSS rules. Therefore, we systematically screened the literature to elaborate on the current status of ontologies applied in CDSS rules, such as rule management, which uses captured CDSS rule usage data and user feedback data to tailor CDSS services to be more accurate, and maintenance, which updates CDSS rules. Through this systematic literature review, we aim to identify the frontiers of ontologies used in CDSS rules. Methods The literature search was focused on the intersection of ontologies; clinical decision support; and rules in PubMed, the Association for Computing Machinery (ACM) Digital Library, and the Nursing & Allied Health Database. Grounded theory and PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) 2020 guidelines were followed. One author initiated the screening and literature review, while 2 authors validated the processes and results independently. The inclusion and exclusion criteria were developed and refined iteratively. Results CDSSs were primarily used to manage chronic conditions, alerts for medication prescriptions, reminders for immunizations and preventive services, diagnoses, and treatment recommendations among 81 included publications. The CDSS rules were presented in Semantic Web Rule Language, Jess, or Jena formats …
Cited by:20
Resources:[HTML] from jmir.org
:view less
A GPT-based Practical Architecture for Conversational Human Digital Twins.
2023
Shahan Mehtab Iqbal, Matias Volonte, Bart P Knijnenburg, Nina C Hubig
AiOfAi@ IJCAI, 33-35
:read more
Authors:Shahan Mehtab Iqbal, Matias Volonte, Bart P Knijnenburg, Nina C Hubig
Publication date:01/01/2023
Journal:AiOfAi@ IJCAI
Pages:33-35
Description:Virtual human twins, also known as digital clones, are computer-generated replicas of human beings that can be used for a variety of purposes, such as entertainment, education, and healthcare. The vision of creating virtual human twins with distinct personalities and emotional responses has the potential to revolutionize these industries by enabling more personalized and engaging experiences for users. This research aims to explore this vision under the aspect of responsponsible artificial intelligence (AI) and identify the tools, technologies, and limitations involved in its realization. We will examine the current state of virtual human twin technology and evaluate the open-source resources available for building them. Additionally, we will investigate the requirements of virtual human twins to be created in a responsible way and analyze the challenges that need to be overcome to create more sophisticated versions. The findings of this research will provide valuable insights into the future of virtual human twin technology and inform future efforts in this field.
Cited by:2
Resources:[PDF] from ceur-ws.org
:view less
Evaluation of breast tumor morphologies from African American and Caucasian patients
2023
A Stone, C Kalahiki, L Li, N Hubig, F Iuricich, H Dunn
Computational and Structural Biotechnology Journal 21, 3459-3465
:read more
Authors:A Stone, C Kalahiki, L Li, N Hubig, F Iuricich, H Dunn
Publication date:01/01/2023
Journal:Computational and Structural Biotechnology Journal
Volume:21
Pages:3459-3465
Publisher:Elsevier
Description:The primary aim of this research was to investigate potential differences of breast tumor morphologies across African American and Caucasian racial groups by utilizing machine learning (ML) and artificial intelligence (AI) methods. While breast cancer disparities can partially be attributed to social determinants of health, tumor biology also contributes to survival outcomes. The rate of breast tumor growth is largely dependent on the extracellular matrix (ECM). Current research suggests the cellular components of the ECM may vary among racial and ethnic populations, and this may contribute to the incidence of cancer in African Americans. We utilized a supervised AI method to evaluate morphological differences between African American and Caucasian breast cancer tumors. Images used for analysis were downloaded from the Cancer Genome Atlas (TCGA) biorepository stored in the NIH Genomic Data Commons …
Cited by:1
Resources:[HTML] from sciencedirect.com
:view less
An end‐to‐end flood stage prediction system using deep neural networks
2023
L Windheuser, R Karanjit, R Pally, S Samadi, NC Hubig
Earth and Space Science 10 (1), e2022EA002385
:read more
Authors:L Windheuser, R Karanjit, R Pally, S Samadi, NC Hubig
Publication date:01/01/2023
Journal:Earth and Space Science
Volume:10
Issue:1
Pages:e2022EA002385
Description:The use of automated methods for detecting and classifying different types of labels in flood images have important applications in hydrologic prediction. In this research, we propose a fully automated end‐to‐end image detection system to predict flood stage data using deep neural networks across two US Geological Survey (USGS) gauging stations, that is, the Columbus and the Sweetwater Creek, Georgia, USA. The images were driven from the USGS live river web cameras, which were strategically located nearby the monitoring stations and refreshed roughly every 30 s. To estimate the flood stage, a U‐Net Convolutional Neural Network (U‐Net CNN) was first stacked on top of a segmentation model for noise and feature reduction that diminished the number of images needed for training. A Long Short‐Term Memory (LSTM), a dense model, and a CNN were then trained to predict the flood stage time series …
Cited by:20
Resources:[PDF] from wiley.com
Full View
:view less
A systematic approach to configuring MetaMap for optimal performance
2022
Xia Jing, Akash Indani, Nina Hubig, Hua Min, Yang Gong, James J Cimino, Dean F Sittig, …
Methods of Information in Medicine 61 (S 02), e51-e63
:read more
Authors:Xia Jing, Akash Indani, Nina Hubig, Hua Min, Yang Gong, James J Cimino, Dean F Sittig, Lior Rennert, David Robinson, Paul Biondich, Adam Wright, Christian Nøhr, Timothy Law, Arild Faxvaag, Ronald Gimbel
Publication date:12/01/2022
Journal:Methods of Information in Medicine
Volume:61
Issue:S 02
Pages:e51-e63
Publisher:Georg Thieme Verlag KG
Description: Background MetaMap is a valuable tool for processing biomedical texts to identify concepts. Although MetaMap is highly configurative, configuration decisions are not straightforward. Objective To develop a systematic, data-driven methodology for configuring MetaMap for optimal performance. Methods MetaMap, the word2vec model, and the phrase model were used to build a pipeline. For unsupervised training, the phrase and word2vec models used abstracts related to clinical decision support as input. During testing, MetaMap was configured with the default option, one behavior option, and two behavior options. For each configuration, cosine and soft cosine similarity scores between identified entities and gold-standard terms were computed for 40 annotated abstracts (422 sentences). The similarity scores were used to calculate and compare the overall percentages of exact matches, similar matches, and …
Cited by:4
Resources:[HTML] from thieme-connect.com
:view less
Interpretable machine learning predicts cardiac resynchronization therapy responses from personalized biochemical and biomechanical features
2022
Anamul Haque, Doug Stubbs, Nina C Hubig, Francis G Spinale, William J Richardson
BMC Medical Informatics and Decision Making 22 (1), 282
:read more
Authors:Anamul Haque, Doug Stubbs, Nina C Hubig, Francis G Spinale, William J Richardson
Publication date:10/31/2022
Journal:BMC Medical Informatics and Decision Making
Volume:22
Issue:1
Pages:282
Publisher:BioMed Central
Description:Background Cardiac Resynchronization Therapy (CRT) is a widely used, device-based therapy for patients with left ventricle (LV) failure. Unfortunately, many patients do not benefit from CRT, so there is potential value in identifying this group of non-responders before CRT implementation. Past studies suggest that predicting CRT response will require diverse variables, including demographic, biomarker, and LV function data. Accordingly, the objective of this study was to integrate diverse variable types into a machine learning algorithm for predicting individual patient responses to CRT. Methods We built an ensemble classification algorithm using previously acquired data from the SMART-AV CRT clinical trial (n = 794 patients). We used five-fold stratified cross-validation on 80% of the patients (n = 635) to train the model with variables collected at 0 months (before initiating CRT), and the remaining 20% of the …
Cited by:6
Resources:[HTML] from springer.com
Full View
:view less
Visualization in Data Science VDS@ KDD 2022
2022
Claudia Plant, Nina C Hubig, Junming Shao, Alvitta Ottley, Liang Gou, Torsten Möller, A…
Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, 4…
:read more
Authors:Claudia Plant, Nina C Hubig, Junming Shao, Alvitta Ottley, Liang Gou, Torsten Möller, Adam Perer, Alexander Lex, Anamaria Crisan
Publication date:08/14/2022
Journal:Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining
Pages:4894-4895
Description:Data science is the practice of deriving insight from data, enabled by modeling, computational methods, interactive visual analysis, and domain-driven problem solving. Data science draws from methodology developed in such fields as applied mathematics, statistics, machine learning, data mining, data management, visualization, and HCI. It drives discoveries in business, economy, biology, medicine, environmental science, the physical sciences, the humanities and social sciences, and beyond. Machine learning and data mining and visualization are integral parts of data science, and essential to enable sophisticated analysis of data. Nevertheless, both research areas are currently still rather separated and investigated by different communities rather independently. The goal of this workshop is to bring researchers from both communities together in order to discuss common interests, to talk about practical issues …
:view less
Application of image processing and convolutional neural networks for flood image classification and semantic segmentation
2022
RJ Pally, S Samadi
Environmental modelling & software 148, 105285
:read more
Authors:RJ Pally, S Samadi
Publication date:02/01/2022
Journal:Environmental modelling & software
Volume:148
Pages:105285
Publisher:Elsevier
Description:Deep learning algorithms are exceptionally valuable tools for collecting and analyzing the catastrophic readiness and countless actionable flood data. Convolutional neural networks (CNNs) are one form of deep learning algorithms widely used in computer vision which can be used to study flood images and assign learnable weights to various objects in the image. Here, we leveraged and discussed how connected vision systems can be used to embed cameras, image processing, CNNs, and data connectivity capabilities for flood label detection. We built a training database service of >9000 images (image annotation service) including the image geolocation information by streaming relevant images from social media platforms, Department of Transportation (DOT) 511 traffic cameras, the US Geological Survey (USGS) live river cameras, and images downloaded from search engines. We then developed a new …
Cited by:90
Resources:[PDF] from sciencedirect.com
:view less
Potentialfinder-Fostering Network Innovation by Connecting Data Owners Using Scaled Business-Relevant Pattern Recognition and Clustering.
2022
Mathias Riechert, Roshan Bhandari, Abhijeet Amle, Abhimanyu Abhinav, Nina C Hubig
HICSS, 1-10
:read more
Authors:Mathias Riechert, Roshan Bhandari, Abhijeet Amle, Abhimanyu Abhinav, Nina C Hubig
Publication date:01/04/2022
Journal:HICSS
Pages:1-10
Description:The advancement in collection, computing and storage technologies has led to an exponential growth of available data in multiple disciplines. However, the human capacity of analyzing this data does not grow at the same rate, leaving a vast amount of potential disparate, invisible and unused. We want to enhance the capability of humans to automatically find relevant patterns in data to leverage potential in this increasing sea of data. We present an innovation network creation framework and Python library that detects exponential growth patterns from publicly available tabular data. It works as a magnifying glass to reveal the most relevant parts of the data and the processes represented by it. The extracted exponential patterns can be useful for topic or disease detection as well as for organisations such as venture capital and consulting firms to improve investment decisions. Additionally, startups and innovation units in corporates can leverage these information to base their business models on insights into sectors, markets or customer segments with exponential growth. To foster the innovation based on the revealed patterns, we connect the respective data owners that uploaded similar patterns. This paper proposes a framework for networked innovation creation including a) an algorithm to automatically detect exponential, b) approaches to scale its application to public tabular data in different sizes and formats, c) a similarity network connecting the found patterns to innovation networks, d) a clustering to group the data owners and enable co-and crowd innovation. We run experiments on large scale data for all steps to provide evidence for cost …
Resources:[PDF] from academia.edu
:view less
Keyphrase Identification with a Limited Labeled Dataset Using Deep Active Learning and Domain Adaptation.
2022
Rohan Goli, Nina C Hubig, Hua Min, Yang Gong, Dean F Sittig, David Robinson, Paul G Bio…
AMIA
:read more
Authors:Rohan Goli, Nina C Hubig, Hua Min, Yang Gong, Dean F Sittig, David Robinson, Paul G Biondich, Adam Wright, Christian Nøhr, Tim Law, Arild Faxvaag, Ronald W Gimbel, Lior Rennert, Xia Jing
Publication date:01/01/2022
Journal:AMIA
:view less
Interpretable machine learning predicts cardiac resynchronization therapy responses from personalized biochemical and biomechanical features
2022
Doug Stubbs, William J Richardson, Anamul Haque, Nina C Hubig, Francis G Spinale
:read more
Authors:Doug Stubbs, William J Richardson, Anamul Haque, Nina C Hubig, Francis G Spinale
Publication date:01/01/2022
Publisher:figshare Academic Research System
Description:Abstract Background Cardiac Resynchronization Therapy (CRT) is a widely used, device-based therapy for patients with left ventricle (LV) failure. Unfortunately, many patients do not benefit from CRT, so there is potential value in identifying this group of non-responders before CRT implementation. Past studies suggest that predicting CRT response will require diverse variables, including demographic, biomarker, and LV function data. Accordingly, the objective of this study was to integrate diverse variable types into a machine learning algorithm for predicting individual patient responses to CRT. Methods We built an ensemble classification algorithm using previously acquired data from the SMART-AV CRT clinical trial (n= 794 patients). We used five-fold stratified cross-validation on 80% of the patients (n= 635) to train the model with variables collected at 0 months (before initiating CRT), and the remaining 20% of the patients (n= 159) were used as a hold-out test set for model validation. To improve model interpretability, we quantified feature importance values using SHapley Additive exPlanations (SHAP) analysis and used Local Interpretable Model-agnostic Explanations (LIME) to explain patient-specific predictions. Results Our classification algorithm incorporated 26 patient demographic and medical history variables, 12 biomarker variables, and 18 LV functional variables, which yielded correct prediction of CRT response in 71% of patients. Additional patient stratification to identify the subgroups with the highest or lowest likelihood of response showed 96% accuracy with 22 correct predictions out of 23 patients in the highest and lowest …
:view less
FlowDyn: a dynamic web-based streamflow prediction pipeline
2021
Nushrat Humaira, S Sadeghi Tabas, S Samadi, Nina C Hubig
2021 IEEE International Conference on Big Data (Big Data), 757-767
:read more
Authors:Nushrat Humaira, S Sadeghi Tabas, S Samadi, Nina C Hubig
Publication date:12/15/2021
Journal:2021 IEEE International Conference on Big Data (Big Data)
Pages:757-767
Publisher:IEEE
Description:Streamflow prediction from meteorological observations provides the basic information for the management of water resources systems. The amount and magnitude of streamflow has long term consequences on human lives and the environment. Various approaches have been proposed to simulate streamflow records, ranging from physically based models to conceptual and fully data-driven models. However, the quality of streamflow prediction can be more reliable with intelligence-based data-driven models that handle nonlinear hydrological processes. We propose FlowDyn, a dynamic web based streamflow prediction pipeline to intelligently simulate and forecast global streamflow time series data. In this research a number of data driven deep learning (DL) models including multi-layer perception (MLP), long short-term memory (LSTM) and a hybrid network of convolutional neural network and LSTM (CNN …
Cited by:4
:view less
Mitigating digital mindlessness
2021
Sushmita Khan, Mehtab Iqbal, Nushrat Humaira, Nina Hubig, Bart Knijnenburg
NA NA (2021), NA
:read more
Authors:Sushmita Khan, Mehtab Iqbal, Nushrat Humaira, Nina Hubig, Bart Knijnenburg
Publication date:08/01/2021
Journal:NA NA (2021), NA
Description:Mindfulness is a well-studied phenomenon and the use of technology to promote mindfulness are quickly gaining popularity. However, in the digital space mindfulness is often overlooked. Instead, there is a rise in deceptive and exploitative techniques in artificial intelligence (AI) that encourage and exploit users’ mindless behaviors. To combat digital mindlessness and the effects of these hazardous AI, this position paper explores the literature in an attempt to define the concept of mindlessness and find ways to reduce it (ie, to promote more mindful behavior). Particularly, mindless-making effects of AI is explored from four thematic viewpoints 1) Filter Bubbles, 2) Dark Patterns, 3) Fake news and
Cited by:3
Resources:[PDF] from ceur-ws.org
:view less
A clinical decision support system (CDSS) ontology to facilitate portable vaccination CDSS rules: preliminary results
2021
Xia Jing, Hua Min, Yang Gong, James J Cimino, David Robinson, Dean F Sittig, Paul Biond…
AMIA Annu Symp Proc
:read more
Authors:Xia Jing, Hua Min, Yang Gong, James J Cimino, David Robinson, Dean F Sittig, Paul Biondich, Adam Wright, Christian Gradhandt Nøhr, Timothy Law, Arild Faxvaag, Akash Indani, Nina Hubig, Ronald Gimbel, Lior Rennert
Publication date:02/21/2021
Journal:AMIA Annu Symp Proc
Description:A clinical decision support system (CDSS) ontology to facilitate portable vaccination CDSS rules: preliminary results - Aalborg University's Research Portal Skip to main navigation Skip to search Skip to main content Aalborg University's Research Portal Home Aalborg University's Research Portal Logo Dansk English Home Profiles Projects Publications Activities Research units Facilities Press/Media Prizes Datasets Impacts Search by keywords, name or affiliation A clinical decision support system (CDSS) ontology to facilitate portable vaccination CDSS rules: preliminary results Xia Jing * , Hua Min, Yang Gong, James J. Cimino, David Robinson, Dean F. Sittig, Paul Biondich, Adam Wright, Christian Gradhandt Nøhr, Timothy Law, Arild Faxvaag, Akash Indani, Nina Hubig, Ronald Gimbel, Lior Rennert * Corresponding author for this work Research output: Contribution to book/anthology/report/conference proceeding › …
Cited by:4
:view less
Flood Informatics: Real Time Flood Level Detection from Multimedia Images and Weather Data
2020
Nushrat Humaira, Nina Hubig, Vidya Samadi
AGU Fall Meeting Abstracts 2020, H140-0008
:read more
Authors:Nushrat Humaira, Nina Hubig, Vidya Samadi
Publication date:12/01/2020
Journal:AGU Fall Meeting Abstracts
Volume:2020
Pages:H140-0008
Description:Successive flooding across the US continues to pose significant challenges for authorities to smartly control the event and promptly communicate vital information to all parties concerned. Detecting roadway segments inundated due to floodwater has important applications for vehicle routing and traffic management decisions. This study introduces a new flood informatics for rapid detection of flood depth and areas across road networks that are crucial factors for real-time transportation system control. We combined crowdsourced data, satellite imagery, National Weather Service real time forecast with USGS data to provide the intelligence for flood detection model. A deep computing system was developed to detect road flooding and compute flood level and inundation areas. The model incrementally updates its prediction with every new input stream. Our method is vastly experimentally evaluated on synthetic and …
:view less
An Intelligent Cyberinfrastructure for IoT Sensing, Flood and Storm water Monitoring and Modeling Systems (Intelligent Riverâ Initiative)
2020
Vidya Samadi, Christopher Post, Calvin Sawyer, Charles Privette III, Nina Hubig
AGU Fall Meeting Abstracts 2020, H177-02
:read more
Authors:Vidya Samadi, Christopher Post, Calvin Sawyer, Charles Privette III, Nina Hubig
Publication date:12/01/2020
Journal:AGU Fall Meeting Abstracts
Volume:2020
Pages:H177-02
Description:Leveraging recent advances in technologies surrounding the Internet of Things," smart" flood and stormwater systems are poised to transform watershed management by enabling ubiquitous real-time sensing and control. Recent applications have demonstrated the potential to improve flood forecasting, stormwater management, and prevent road flooding and sewer overflow. However, adoption of smart flood and stormwater systems has been hindered by a limited number of successful case studies, along with a lack of guidance on how smart flood and stormwater systems should be designed and built. This research introduces Intelligent Riverâ Initiative—an open-source, end-to-end platform for real-time monitoring, modeling and control of a rapid response watershed system in South Carolina. Intelligent Riverâ Initiative includes (i) a field-deployed sensor fabric and uplink infrastructure,(ii) real-time data streaming …
:view less
Human-centric preference modeling for virtual agents
2020
Bart P Knijnenburg, Nina Hubig
Proceedings of the 20th ACM International Conference on Intelligent Virtual Agents, 1-3
:read more
Authors:Bart P Knijnenburg, Nina Hubig
Publication date:10/20/2020
Journal:Proceedings of the 20th ACM International Conference on Intelligent Virtual Agents
Pages:1-3
Description:Intelligent virtual agents increasingly use machine learning to model the needs and preferences of their users in order to give personalized support. In this paper we present a vision of "human-centric modeling" that moves this practice beyond the automated estimation of users' current preferences based on observable behaviors to a modeling process that a) considers users' hidden characteristics rather than observable behaviors, b) models possible futures rather than users' current state, and c) gives users an opportunity to interactively control this modeling process. This vision focuses on the long-term needs of individual humans and a genuine betterment of society.
Cited by:4
:view less
Hyperinsight: Data exploration deep inside hyper
2017
Nina Hubig, Linnea Passing, Maximilian E Schüle, Dimitri Vorona, Alfons Kemper, Thomas …
Proceedings of the 2017 ACM on Conference on Information and Knowledge Management, 2467…
:read more
Authors:Nina Hubig, Linnea Passing, Maximilian E Schüle, Dimitri Vorona, Alfons Kemper, Thomas Neumann
Publication date:11/06/2017
Journal:Proceedings of the 2017 ACM on Conference on Information and Knowledge Management
Pages:2467-2470
Description:Nowadays we are drowning in data of various varieties. For all these mixed types and categories of data there exist even more different analysis approaches, often done in single hand-written solutions. We propose to extend HyPer, a main memory database system to a uniform data agent platform following the one system fits all approach for solving a wide variety of data analysis problems. We achieve this by applying a flexible operator concept to a set of various important data exploration algorithms. With that, HyPer solves analytical questions using clustering, classification, association rule mining and graph mining besides standard HTAP (Hybrid Transaction and Analytical Processing) workloads on the same database state. It enables to approach the full variety and volume of HTAP extended for data exploration (HTAPx), and only needs knowledge of already introduced SQL extensions that are automatically …
Cited by:17
Resources:[PDF] from tum.de
:view less
Information-theoretic non-redundant subspace clustering
2017
Nina Hubig, Claudia Plant
Pacific-Asia Conference on Knowledge Discovery and Data Mining, 198-209
:read more
Authors:Nina Hubig, Claudia Plant
Publication date:04/23/2017
Journal:Pacific-Asia Conference on Knowledge Discovery and Data Mining
Pages:198-209
Publisher:Springer International Publishing
Description:A comprehensive understanding of complex data requires multiple different views. Subspace clustering methods open up multiple interesting views since they support data objects to be assigned to different clusters in different subspaces. Conventional subspace clustering methods yield many redundant clusters or control redundancy by difficult to set parameters. In this paper, we employ concepts from information theory to naturally trade-off the two major properties of a subspace cluster: The quality of a cluster and its redundancy with respect to the other clusters. Our novel algorithm NORD (for NOn-ReDundant) efficiently discovers the truly relevant clusters in complex data sets without requiring any kind of threshold on their redundancy. NORD also exploits the concept of microclusters to support the detection of arbitrarily-shaped clusters. Our comprehensive experimental evaluation shows the …
Cited by:4
:view less
SQL-and Operator-centric Data Analytics in Relational Main-Memory Databases.
2017
Linnea Passing, Manuel Then, Nina C Hubig, Harald Lang, Michael Schreier, Stephan Günne…
EDBT, 84-95
:read more
Authors:Linnea Passing, Manuel Then, Nina C Hubig, Harald Lang, Michael Schreier, Stephan Günnemann, Alfons Kemper, Thomas Neumann
Publication date:03/21/2017
Journal:EDBT
Pages:84-95
Description:Data volume and complexity continue to increase, as does the need for insight into data. Today, data management and data analytics are most often conducted in separate systems: database systems and dedicated analytics systems. This separation leads to time-and resource-consuming data transfer, stale data, and complex IT architectures. In this paper we show that relational main-memory database systems are capable of executing analytical algorithms in a fully transactional environment while still exceeding performance of state-of-the-art analytical systems rendering the division of data management and data analytics unnecessary. We classify and assess multiple ways of integrating data analytics in database systems. Based on this assessment, we extend SQL with a non-appending iteration construct that provides an important building block for analytical algorithms while retaining the high expressiveness of the original language. Furthermore, we propose the integration of analytics operators directly into the database core, where algorithms can be highly tuned for modern hardware. These operators can be parameterized with our novel user-defined lambda expressions. As we integrate lambda expressions into SQL instead of proposing a new proprietary query language, we ensure usability for diverse groups of users. Additionally, we carry out an extensive experimental evaluation of our approaches in HyPer, our full-fledged SQL main-memory database system, and show their superior performance in comparison to dedicated solutions.
Cited by:66
Resources:[PDF] from tum.de
:view less
Analyzing and Predicting Large Vector-, Graph-and Spatio-Temporal Data
2017
Nina Hubig
:read more
Authors:Nina Hubig
Publication date:01/01/2017
Description:Large social graph datasets, complex vector data and petabytes of sensor data are being generated every day. Employing this flood of data for the benefit of all, is one of the main challenges of the 21st century. This thesis advances the field of data mining and machine learning for a variety of data types. Additionally it shows a way to integrate these methods in modern main-memory databases, employing examples of classical clustering and classification approaches.
:view less
Detection and prediction of natural hazards using large-scale environmental data
2017
Nina Hubig, Philip Fengler, Andreas Züfle, Ruixin Yang, Stephan Günnemann
Advances in Spatial and Temporal Databases: 15th International Symposium, SSTD 2017, Ar…
:read more
Authors:Nina Hubig, Philip Fengler, Andreas Züfle, Ruixin Yang, Stephan Günnemann
Publication date:01/01/2017
Journal:Advances in Spatial and Temporal Databases: 15th International Symposium, SSTD 2017, Arlington, VA, USA, August 21–23, 2017, Proceedings 15
Pages:300-316
Publisher:Springer International Publishing
Description:Recent developments in remote sensing have made it possible to instrument and sense the physical world with high resolution and fidelity. Consequently, very large spatio-temporal environmental data sets, have become available to the research community. Such data consists of time-series, starting as early as 1973, monitoring up to thousands of environmental parameters, for each spatial region of a resolution as low as . To make this flood of data actionable, in this work, we employ a data driven approach to detect and predict natural hazards. Our supervised learning approach learns from labeled historic events. We describe each event by a three-mode tensor, covering space, time and environmental parameters. Due to the very large number of environmental parameters, and the possibility of latent features hidden within these parameters, we employ a tensor factorization approach to learn …
Cited by:12
:view less
Generalized independent subspace clustering
2016
Wei Ye, Samuel Maurus, Nina Hubig, Claudia Plant
2016 IEEE 16th international conference on data mining (ICDM), 569-578
:read more
Authors:Wei Ye, Samuel Maurus, Nina Hubig, Claudia Plant
Publication date:12/12/2016
Journal:2016 IEEE 16th international conference on data mining (ICDM)
Pages:569-578
Publisher:IEEE
Description:Data can encapsulate different object groupings in subspaces of arbitrary dimension and orientation. Finding such subspaces and the groupings within them is the goal of generalized subspace clustering. In this work we present a generalized subspace clustering technique capable of finding multiple non-redundant clusterings in arbitrarily-oriented subspaces. We use Independent Subspace Analysis (ISA) to find the subspace collection that minimizes the statistical dependency (redundancy) between clusterings. We then cluster in the arbitrarily-oriented subspaces identified by ISA. Our algorithm ISAAC (Independent Subspace Analysis and Clustering) uses the Minimum Description Length principle to automatically choose parameters that are otherwise difficult to set. We comprehensively demonstrate the effectiveness of our approach on synthetic and real-world data.
Cited by:35
:view less
Effiziente Integration von Data-und Graph-Mining-Algorithmen in relationale Datenbanksysteme.
2015
Manuel Then, Linnea Passing, Nina C Hubig, Stephan Günnemann, Alfons Kemper, Thomas Neu…
LWA, 45-49
:read more
Authors:Manuel Then, Linnea Passing, Nina C Hubig, Stephan Günnemann, Alfons Kemper, Thomas Neumann
Publication date:01/01/2015
Journal:LWA
Pages:45-49
Description:Big Data “immer mehr in das Zentrum der Aufmerksamkeit. Um diese komplexen Datenanalysen effizient zu ermöglichen liegt es nahe, sie in die am weitesten verbreiteten Datenspeicher zu integrieren–in relationale Datenbanksysteme. Dies führt zu interessanten Fragestellungen nicht nur im Bereich der technischen Integration sondern besonders auch in der Anfragespezifikation und-auswertung. In diesem Kurzbeitrag beschreiben wir, wie Algorithmen zur Datenanalyse effizient und nutzerfreundlich in das relationale Hauptspeicherdatenbanksystem HyPer integriert werden können. Wir evaluieren unseren Ansatz anhand eines Vergleichs mit zwei verbreiteten Datenanalysesystemen auf Graph-und Vektordaten.
Cited by:7
Resources:[PDF] from ceur-ws.org
:view less
Musical similarity analysis based on chroma features and text retrieval methods
2015
David Englmeier, Nina Hubig, Sebastian Goebl, Christian Böhm
Datenbanksysteme für Business, Technologie und Web (BTW 2015)-Workshopband, 183-192
:read more
Authors:David Englmeier, Nina Hubig, Sebastian Goebl, Christian Böhm
Publication date:01/01/2015
Journal:Datenbanksysteme für Business, Technologie und Web (BTW 2015)-Workshopband
Pages:183-192
Publisher:Gesellschaft für Informatik eV
Description:At the present day the world wide web is full of music. Highly effective algorithms for music compression and high data storage has made it easy to access all kind of music easily. However, it is not possible to look for a similar piece of music or a sound as easily as to google for a similar kind of text. Music is filtered by its title or artist. Although musicians can publish their compositions in a second, they will only be found by high youtube ratings or by market basket analysis. Less known artists need much luck to get heard, although their music might just be what people want to hear. To approach this issue, we propose a new framework called MIRA (Music Information Retrieval Application) for analyzing audio files with existing Information Retrieval (IR) methods. Text retrieval has already yielded many highly efficient and generally accepted methods to assess the semantic distance of different text. We use these methods by translating music into equivalent audio words based on chroma features. We show that our framework can easily match music interpreted even by different artists.
Cited by:4
Resources:[PDF] from gi.de
:view less
Development and evaluation of a facebook-based product advisor for online dating sites
2015
Martin Winter, Sebastian Goebl, Nina Hubig, Christopher Pleines, Christian Böhm
Datenbanksysteme für Business, Technologie und Web (BTW 2015)-Workshopband, 213-222
:read more
Authors:Martin Winter, Sebastian Goebl, Nina Hubig, Christopher Pleines, Christian Böhm
Publication date:01/01/2015
Journal:Datenbanksysteme für Business, Technologie und Web (BTW 2015)-Workshopband
Pages:213-222
Publisher:Gesellschaft für Informatik eV
Description:Social networks play an important role in Web 2.0. For many users establishing contacts and staying in touch in the virtual world is more than just a spare time filler. In social networks like Facebook they provide much information about themselves in user profiles. Also for online dating the focus is on establishing new contacts. In general, three types of dating sites can be distinguished: more serious dating agencies, less focused singles' platforms, and casual dating sites. In the proposed paper we develop a product advisor that uses the Facebook profile information provided by a user to classify her or him to one of the three dating site categories which is most suitable for the user's purpose. For classification we use Naive Bayes. To train the classifier we investigate the correlation between profile information and choice of dating sites. We also evaluate this correlation on collected data of representative German dating sites. Although a sharp distinction is hard to find, tendencies and enlightening insights are revealed by the collected data.
Resources:[PDF] from gi.de
:view less
Centroid clustering of cellular lineage trees
2014
Valeriy Khakhutskyy, Michael Schwarzfischer, Nina Hubig, Claudia Plant, Carsten Marr, M…
Information Technology in Bio-and Medical Informatics: 5th International Conference, IT…
:read more
Authors:Valeriy Khakhutskyy, Michael Schwarzfischer, Nina Hubig, Claudia Plant, Carsten Marr, Michael A Rieger, Timm Schroeder, Fabian J Theis
Publication date:01/01/2014
Journal:Information Technology in Bio-and Medical Informatics: 5th International Conference, ITBAM 2014, Munich, Germany, September 2, 2014. Proceedings 5
Pages:15-29
Publisher:Springer International Publishing
Description:Trees representing hierarchical knowledge are prevalent in biology and medicine. Some examples are phylogenetic trees, the hierarchical structure of biological tissues and cell lines. The increasing throughput of techniques generating such trees poses new challenges to the analysis of tree ensembles. Some typical tasks include the determination of common patterns of lineage decisions in cellular differentiation trees. Partitioning the dataset is crucial for further analysis of the cellular genealogies. In this work, we develop a method to cluster labeled binary tree structures. Furthermore, for every cluster our method selects a centroid tree that captures the characteristic mitosis patterns of the group. We evaluate this technique on synthetic data and apply it to experimental trees that embody the lineages of differentiating cells under specific conditions over time. The results of the cell lineage trees are thoroughly …
Cited by:3
:view less
Monitoring probabilistic threshold SUM query processing in uncertain streams
2014
Nina Hubig, Andreas Züfle, Tobias Emrich, Matthias Renz, Mario A Nascimento, Hans-Peter…
Database Systems for Advanced Applications: 19th International Conference, DASFAA 2014,…
:read more
Authors:Nina Hubig, Andreas Züfle, Tobias Emrich, Matthias Renz, Mario A Nascimento, Hans-Peter Kriegel
Publication date:01/01/2014
Journal:Database Systems for Advanced Applications: 19th International Conference, DASFAA 2014, Bali, Indonesia, April 21-24, 2014. Proceedings, Part I 19
Pages:420-435
Publisher:Springer International Publishing
Description:Many sources of data streams, e.g. geo-spatial streams derived from GPS-tracking systems or sensor streams provided by sensor networks are inherently uncertain due to impreciseness of sensing devices, due to outdated information, and due to human errors. In order to support data analysis on such data, aggregation queries are an important class of queries. This paper introduces a scalable approach for continuous probabilistic SUM query processing in uncertain stream environments. Here we consider an uncertain stream as a stream of uncertain values, each given by a probability distribution among the domain of the sensor values. Continuous probabilistic sum queries maintain the probability distribution of the sum of possible sensor values actually derived from the streaming environment. Our approach is able to efficiently compute the probabilistic SUM according to the possible world semantics, i.e …
Cited by:1
:view less
Active density-based clustering
2013
Son T Mai, Xiao He, Nina Hubig, Claudia Plant, Christian Böhm
2013 IEEE 13th International Conference on Data Mining, 508-517
:read more
Authors:Son T Mai, Xiao He, Nina Hubig, Claudia Plant, Christian Böhm
Publication date:12/07/2013
Journal:2013 IEEE 13th International Conference on Data Mining
Pages:508-517
Publisher:IEEE
Description:The density-based clustering algorithm DBSCAN is a fundamental technique for data clustering with many attractive properties and applications. However, DBSCAN requires specifying all pair wise (dis)similarities among objects that can be non-trivial to obtain in many applications. To tackle this problem, in this paper, we propose a novel active density-based clustering algorithm, named Act-DBSCAN, which works under a restricted number of used pair wise similarities. Act-DBSCAN exploits the pair wise lower-bounding (LB) similarities to initialize the cluster structure. Then, it adaptively selects the most informative pair wise LB similarities to update with the real ones in order to reconstruct the result until the budget limitation is reached. The goal is to approximate as much as possible the true clustering result with each update. Our Act-DBSCAN framework is built upon a proposed probabilistic model to score the …
Cited by:44
:view less
Compression-based graph mining exploiting structure primitives
2013
Jing Feng, Xiao He, Nina Hubig, Christian Böhm, Claudia Plant
2013 IEEE 13th International Conference on Data Mining, 181-190
:read more
Authors:Jing Feng, Xiao He, Nina Hubig, Christian Böhm, Claudia Plant
Publication date:12/07/2013
Journal:2013 IEEE 13th International Conference on Data Mining
Pages:181-190
Publisher:IEEE
Description:How can we retrieve information from sparse graphs? Traditional graph mining approaches focus on discovering dense patterns inside complex networks, for example modularity-based or cut-based methods. However, most real world data sets are very sparse. Nevertheless, traditional approaches tend to omit interesting sparse patterns like stars. In this paper, we propose a novel graph mining technique modeling the transitivity and the hub ness of a graph using structure primitives. We exploit these structure primitives for effective graph compression using the Minimum Description Length Principle. The compression rate is an unbiased measure for the transitivity or hub ness and therefore provides interesting insights into the structure of even very sparse graphs. Since real graphs can be composed of sub graphs of different structures, we propose a novel algorithm CXprime (Compression-based exploiting …
Cited by:17
:view less
Continuous probabilistic sum queries in wireless sensor networks with ranges
2012
Nina Hubig, Andreas Züfle, Tobias Emrich, Mario A Nascimento, Matthias Renz, Hans-Peter…
Scientific and Statistical Database Management: 24th International Conference, SSDBM 20…
:read more
Authors:Nina Hubig, Andreas Züfle, Tobias Emrich, Mario A Nascimento, Matthias Renz, Hans-Peter Kriegel
Publication date:01/01/2012
Journal:Scientific and Statistical Database Management: 24th International Conference, SSDBM 2012, Chania, Crete, Greece, June 25-27, 2012. Proceedings 24
Pages:96-105
Publisher:Springer Berlin Heidelberg
Description:Data measured in wireless sensor networks are inherently imprecise. Aggregate queries are often used to analyze the collected data in order to alleviate the impact of such imprecision. In this paper we will deal with the imprecision in the measured values explicitly by employing a probabilistic approach and we focus on one particular type of aggregate query, namely the SUM query.
Cited by:3
Resources:[PDF] from uni-muenchen.de
:view less
Natural Language Processing and Social Determinants of Health in Mental Health Research: An Artificial Intelligence Assisted Scoping Review
Dmitry Scherbakov, Nina Hubig, Leslie Andrew Lenert, Alexander V Alekseyenko, Jihad S O…
:read more
Authors:Dmitry Scherbakov, Nina Hubig, Leslie Andrew Lenert, Alexander V Alekseyenko, Jihad S Obeid
Description:Background: The usage of natural language processing (NLP) in mental health research is increasing with a wide range of applications and datasets being investigated. Objective: This review aims to summarize the usage NLP in mental health research, with a special focus on the types of text datasets and the usage of social determinants of health (SDOH) in NLP projects related to mental health. Methods: The search was conducted in September 2024 using a broad search strategy in PubMed, Scopus, and CINAHL Complete. All citations were uploaded to Covidence online software. The screening and extraction process took place in Covidence with the help of a custom large language model (LLM) module developed by our team. This LLM module was calibrated and tuned to substitute human reviewers. Results: The screening process, assisted by the custom LLM, led to the inclusion of 1,768 studies in the final review. The majority of the reviewed studies (n= 665, 42.8%) utilized clinical data as their primary text dataset, followed by social media datasets (n= 523, 33.7%). The United States contributed the highest number of studies (n= 568, 36.6%), with depression (n= 438, 28.2%) and suicide (n= 240, 15.5%) being the most frequently investigated mental health issues. Traditional demographic variables such as age (n= 877, 56.5%) and gender (n= 760, 49.0%) were commonly extracted, while SDOH factors were less frequently reported, with urban/rural status being the most used (n= 19, 1.2%). Over half of the citations (n= 826, 53.2%) did not provide clear information on dataset accessibility, although a sizable number of studies (n= 304, 19.6 …
Resources:[PDF] from amazonaws.com
:view less
Automated analysis of neuronal morphologies via python and (collaborative) open source databases
Nina Hubig, Hans J Ohlbach, Thomas Wachtler, Philipp L Rautenberg
:read more
Authors:Nina Hubig, Hans J Ohlbach, Thomas Wachtler, Philipp L Rautenberg
Description:Neuronal morphology plays a fundamental role for the information processing capabilities of neurons. To fully exploit the information gained by neuron reconstructions, it is crucial for neuroscientific researchers to analyze a large amount of data in a fast, effective, and automated way. We developed a new system for the analysis of morphological data that is based on a high-level object-oriented programming language that was integrated into an object-relational database management system. This approach enables to connect automatized data analysis procedures directly to the stored data. In our (experimental) system we used Python as object oriented language which is commonly used in the neuroscientific community. It is a versatile language that can be easily applied to solve several tasks that research laboratories face almost every day: data manipulation, biological data retrieval and parsing, automation, or …
:view less
:back to overview