Review
Biochemical Research Methods
Anna Bernasconi, Arif Canakoglu, Marco Masseroli, Pietro Pinoli, Stefano Ceri
Summary: With the outbreak of COVID-19, the research community is making unprecedented efforts to understand and mitigate the impact of the pandemic. Organizations are focusing on COVID-19 virus data and services, while new organizations and resources are emerging to lay the groundwork for future pandemic studies.
BRIEFINGS IN BIOINFORMATICS
(2021)
Article
Computer Science, Information Systems
Umer Rashid, Khalid Saleem, Adeel Ahmed
Summary: The core issues in multimedia content exploration lie in the limitations of linear interaction mechanisms and the underutilization of various information modalities. This research proposes a nonlinear and multimodal approach for exploring multimedia document results, utilizing result spaces and SUI design. Experimental results show high user satisfaction and utilization of SUI components.
MULTIMEDIA TOOLS AND APPLICATIONS
(2021)
Article
Environmental Sciences
Ning Zhang, Francesco Nex, George Vosselman, Norman Kerle
Summary: This research focuses on using deep learning to detect victims in disaster debris, proposes a method to generate harmonious composite images for training, and significantly improves detection accuracy.
Article
Computer Science, Information Systems
Dan Li, Tong Xu, Peilun Zhou, Weidong He, Yanbin Hao, Yi Zheng, Enhong Chen
Summary: This article proposes a social context-aware framework that fuses visual and social contexts to profile persons in more semantic perspectives and achieves better performance in handling person search task in complex scenarios. The framework segments videos into independent scene units, abstracts social contexts, constructs inner-personal links through a graph formulation operation, and performs relation-aware label propagation to identify characters' occurrences. Experiments demonstrate that the proposed solution outperforms competitive baselines on real-world datasets.
ACM TRANSACTIONS ON INFORMATION SYSTEMS
(2022)
Review
Pharmacology & Pharmacy
Todd Millecam, Austin J. Jarrett, Naomi Young, Dana E. Vanderwall, Dennis Della Corte
Summary: The Allotrope Foundation is a group of pharmaceutical, device vendor, and software companies that develops technologies to simplify the exchange of electronic data. This article presents the history, structure, members and partners of the AF, as well as an overview of the technologies. Insights into the adoption and development of the technologies are provided through a summary of the Fall 2020 Allotrope Connect virtual conference.
DRUG DISCOVERY TODAY
(2021)
Article
Engineering, Electrical & Electronic
Zhong Xiang, Yujia Shen, Miao Ma, Miao Qian
Summary: This article presents an advanced fabric defect detection model based on deep learning algorithm. The model incorporates a parallel dilated attention module (PDAM) and a hook-shaped feature pyramid network (HookFPN), which leads to improved detection efficiency without sacrificing accuracy.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2023)
Article
Computer Science, Theory & Methods
Rafael Padilha, Tawfiq Salem, Scott Workman, Fernanda A. Andalo, Anderson Rocha, Nathan Jacobs
Summary: This study presents the problem of detecting timestamp manipulation and proposes an end-to-end approach using supervised consistency verification to verify the consistency of image capture time, content, and geographic location. Through experiments, the method shows significant improvement in classification accuracy on a large dataset and demonstrates the ability to estimate possible capture time in scenarios with missing metadata.
IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY
(2022)
Article
Chemistry, Multidisciplinary
Damjana Drobne
Summary: This paper proposes a metadata list for reporting nanotoxicity experiments to add toxicological context to studies. Existing standards focus on experimental conditions, material characteristics, and specific experiments, but lack interpretation of aims and outcomes of nanotoxicity studies. The proposed checklist aims to enhance reporting and integration of NM toxicity data and knowledge.
Article
Computer Science, Interdisciplinary Applications
Cheng Peng, Tong Ge, Yunhai Wang
Summary: This paper examines the problem of visualizing word search results in documents and proposes two design goals: minimizing user interactions and optimizing the onscreen information of search results. Two visualization techniques, a structured list and full sentence snippets, are introduced and compared with commonly used visual interfaces through a user study. The results show that the structured list is favored by users and leads to more efficient search.
JOURNAL OF VISUALIZATION
(2023)
Article
Information Science & Library Science
Shaobo Liang, Linfeng Yu
Summary: This paper aims to explore users' voice search behavior in human-vehicle interaction. Mixed research methods, including questionnaires and interviews, were employed to study users' voice search content, search need, search motivation, satisfaction, barriers, and suggestions. The findings have practical implications for optimizing the voice search interaction system and improving the service in the context of human-vehicle interaction.
Article
Computer Science, Information Systems
Riccardo Albertoni, Monica De Martino, Alfonso Quarati
Summary: The paper discusses evaluating and documenting the quality of controlled vocabularies to promote comparison and enhance semantic interoperability. Using the Analytical Hierarchy Process (AHP), overall quality and ranks of controlled vocabularies can be assessed by integrating different quality dimensions based on decision maker's needs. Selecting a set of e-Government controlled vocabularies as a testbed and providing updated quality values as Linked Data can facilitate interoperability while multi-step guidelines for W3C recommendations can ensure machine-readable quality metadata, promoting reliability and re-usability.
IEEE TRANSACTIONS ON EMERGING TOPICS IN COMPUTING
(2021)
Article
Computer Science, Information Systems
Alfan Farizki Wicaksono, Alistair Moffat
Summary: This paper introduces a session-based offline evaluation framework for measuring the overall usefulness of search sessions. By modeling data from two commercial search engines, the user conditional continuation probability and user conditional reformulation probability are proposed to develop new metrics that show greater correlation with observed user behavior during search sessions.
INFORMATION PROCESSING & MANAGEMENT
(2021)
Article
Computer Science, Information Systems
Anubhav Shivhare, Vishal Krishna Singh, Manish Kumar
Summary: This study improves the existing clustering methods and event detection in IoT by dividing the region into sub-regions based on user context parameters to achieve accurate, context-aware, and discriminatory event detection, and proposes energy-efficient data transmission. Simulation experiments show that the proposed scheme outperforms the existing schemes in terms of detection accuracy and network lifetime.
Article
Computer Science, Artificial Intelligence
Faizan Ahmad, Ahmed Abbasi, Brent Kitchens, Donald Adjeroh, Daniel Zeng
Summary: Adverse event detection is crucial for various real-world applications. This paper proposes a novel deep learning framework called DeepSAVE, which utilizes user search query logs to detect adverse events. Experimental results demonstrate that DeepSAVE outperforms existing detection methods, and each component of DeepSAVE significantly contributes to its overall performance.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2022)
Article
Engineering, Electrical & Electronic
Yongping Zhai, Junhua Wang, Jinsheng Deng, Guanghui Yue, Wei Zhang, Chang Tang
Summary: This study introduces a global context guided hierarchically residual feature refinement network for defocus blur detection, which improves the final detection performance by aggregating different feature information and utilizing methods such as multi-scale dilation convolution. Extensive experiments validate the effectiveness of the proposed network compared to other state-of-the-art methods in terms of both efficiency and accuracy.
Review
Computer Science, Artificial Intelligence
Samuel Sousa, Roman Kern
Summary: This article systematically reviews over sixty DL methods for privacy-preserving NLP published between 2016 and 2020, covering classification, privacy threats, metrics, and challenges in real-world scenarios.
ARTIFICIAL INTELLIGENCE REVIEW
(2023)
Article
Engineering, Biomedical
Changqing Lu, Shreyasi Pathak, Gwenn Englebienne, Christin Seifert
Summary: Machine learning based sleep scoring methods aim to automate the process of annotating polysomnograms with sleep stages. However, most multi-channel multi-modal models in the literature showed little performance improvement compared to single-channel EEG models. In this paper, we investigate the specific features in single-channel EEG models that contribute to their high performance and analyze the extent to which multi-channel multi-modal models utilize information from different channels. Our findings suggest that incorporating advanced methods for aggregating channel information may improve sleep scoring performance for multi-channel multi-modal models.
IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING
(2023)
Article
Engineering, Manufacturing
J. G. Hoffer, B. C. Geiger, R. Kern
Summary: This research presents a method that combines stacked Gaussian processes (stacked GP) with target vector Bayesian optimization (BO) to solve multi-objective inverse problems of chained manufacturing processes. Through stacked GPs, epistemic uncertainty can be propagated through all chained manufacturing processes. In the optimization of chained processes, there are options to use a single unified surrogate model for the entire joint chain or have a surrogate model for each individual process and cascade the optimization from the last to the first process. For improved target vector BO results of chained processes, the proposed approach combines methods that under- or overestimate uncertainties in an ensemble for rank aggregation. Thorough analysis and evaluation are conducted on artificial use cases and a typical manufacturing process chain.
CIRP JOURNAL OF MANUFACTURING SCIENCE AND TECHNOLOGY
(2023)
Review
Computer Science, Theory & Methods
Meike Nauta, Jan Trienes, Shreyasi Pathak, Elisa Nguyen, Michelle Peters, Yasmin Schmitt, Joerg Schloetterer, Maurice Van Keulen, Christin Seifert
Summary: The evaluation of explanations for machine learning models is a complex concept that should not be solely based on subjective validation. This study identifies 12 conceptual properties that should be considered for a comprehensive assessment of explanation quality. The evaluation practices of over 300 papers introducing explainable artificial intelligence (XAI) methods in the past 7 years were systematically reviewed, finding that one-third of the papers exclusively relied on anecdotal evidence and one-fifth evaluated with users. The study also provides an extensive overview of quantitative XAI evaluation methods, offering researchers and practitioners concrete tools for validation and benchmarking.
ACM COMPUTING SURVEYS
(2023)
Review
Computer Science, Information Systems
Philipp Gabler, Bernhard C. Geiger, Barbara Schuppler, Roman Kern
Summary: Superficially, read and spontaneous speech are two main types of training data in automatic speech recognition, but they are fundamentally different due to the way the audio signal is generated. This review introduces causal reasoning into automatic speech recognition, highlighting the impact of data generation processes on inference and performance. By applying a causal perspective, this work discusses the relationship between data generation mechanisms, learning, and prediction in speech data. Furthermore, the authors argue that a causal perspective can enhance the understanding of models in speech processing.
Article
Computer Science, Information Systems
Michael Granitzer, Stefan Voigt, Noor Afshan Fathima, Martin Golasowski, Christian Guetl, Tobias Hecking, Gijs Hendriksen, Djoerd Hiemstra, Jan Martinovic, Jelena Mitrovic, Izidor Mlakar, Stavros Moiras, Alexander Nussbaumer, Per oester, Martin Potthast, Marjana Sencar Srdic, Sharikadze Megi, Katerina Slaninova, Benno Stein, Arjen P. de Vries, Vit Vondrak, Andreas Wagner, Saber Zerhoudi
Summary: Web search is crucial for the digital economy, but the dominance of a few gatekeepers has resulted in a closed ecosystem and potential quality sacrifice. To encourage openness and choice, an Open Web Index (OWI) is proposed, based on open data principles, legal compliance, and collaborative technology development. The combination of an open index with declarative search engines will facilitate the development of vertical search engines and innovative web data products, creating a fair and open information space. The EU-funded project OpenWebSearch.EU marks the first step towards realizing an Open Web Index.
JOURNAL OF THE ASSOCIATION FOR INFORMATION SCIENCE AND TECHNOLOGY
(2023)
Article
Computer Science, Artificial Intelligence
Luca Malinverno, Vesna Barros, Francesco Ghisoni, Giovanni Visona, Roman Kern, Philip J. Nickel, Barbara Elvira Ventura, Ilija Simic, Sarah Stryeck, Francesca Manni, Cesar Ferri, Claire Jean-Quartier, Laura Genga, Gabriele Schweikert, Mario Lovri, Michal Rosen-Zvi
Summary: This study aimed to analyze the association between COVID-19 and the advancement of explainable artificial intelligence (XAI) research. By extracting relevant studies from the PubMed database and manual labeling, the study found that the emergence of COVID-19 may have driven the attention towards XAI and accelerated its development trends.
Review
Computer Science, Information Systems
Shafaq Siddiqi, Faiza Qureshi, Stefanie Lindstaedt, Roman Kern
Summary: This study aims to systematically review outlier detection methods for non-IID data published between 2015 and 2023. The study focuses on data characteristics, methods, and evaluation measures. It provides a comprehensive overview of the characteristics of non-IID data, recent methods for outlier detection, and evaluation metrics. A taxonomy is proposed for organizing these methods and open challenges in outlier detection for non-IID are discussed.
Proceedings Paper
Computer Science, Artificial Intelligence
Lorenz Wendlinger, Michael Granitzer, Christofer Fellicious
Summary: Neural Architecture Search can find high-performance task specific neural network architectures. Using surrogate models as performance predictors can reduce the need for costly evaluations. Our deep graph learning approach achieves state-of-the-art performance in multiple NAS performance prediction benchmarks.
MACHINE LEARNING, OPTIMIZATION, AND DATA SCIENCE, LOD 2022, PT II
(2023)
Proceedings Paper
Computer Science, Artificial Intelligence
Victor Adriel de Jesus Oliveira, Gernot Rottermanner, Magdalena Boucher, Stefanie Groessbacher, Peter Judmaier, Werner Bailer, Georg Thallinger, Thomas Kurz, Jakob Frank, Christoph Bauer, Gabriele Froeschl, Michael Batlogg
Summary: While AI-based audiovisual analysis tools have made significant progress, integrating them into media production and archiving workflows remains challenging due to mismatched annotations, lack of alignment with existing workflows, and the need for accurate metadata. We propose a system that utilizes AI-based analysis methods for annotation and search in media archive applications. To enhance communication and gather feedback, Taylor, an artificial intelligence agent, is included in the user interface.
MULTIMEDIA MODELING, MMM 2023, PT II
(2023)
Proceedings Paper
Computer Science, Artificial Intelligence
Werner Bailer, Hannes Fassold
Summary: To support annotation tasks in visual media production and archiving, two datasets, People@Places and ToDY, are proposed. These datasets cover the annotation of scene bustle, shot cinematographic type, and shot time of day and season. Automatic annotations are created using a toolchain and manually verified and corrected. Baseline results using the EfficientNet-B3 model pretrained on Places365 dataset are provided.
MULTIMEDIA MODELING, MMM 2023, PT I
(2023)
Article
Computer Science, Artificial Intelligence
Maximilian B. Toller, Bernhard C. Geiger, Roman Kern
Summary: Rate-distortion theory-based outlier detection utilizes good data compression to encode outliers with unique symbols. We propose Cluster Purging as an extension of clustering-based outlier detection, allowing the assessment of clustering representivity and the identification of data best represented by individual unique clusters. We present two efficient algorithms for Cluster Purging, one parameter-free and the other allowing tuning in supervised setups.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2023)
Proceedings Paper
Computer Science, Artificial Intelligence
Werner Bailer, Rahel Arnold, Vera Benz, Davide Alessandro Coccomini, Anastasios Gkagkas, Gylfi Thor Gudmundsson, Silvan Heller, Bjorn Thor Jonsson, Jakub Lokoc, Nicola Messina, Nick Pantelidis, Jiaxin Wu
Summary: This paper proposes a process for refining known-item and open-set type queries in interactive video retrieval evaluations. It emphasizes the importance of proper task interpretations.
PROCEEDINGS OF THE 2023 ACM INTERNATIONAL CONFERENCE ON MULTIMEDIA RETRIEVAL, ICMR 2023
(2023)
Proceedings Paper
Computer Science, Artificial Intelligence
Cathal Gurrin, Liting Zhou, Graham Healy, Bjorn Thornor Jonsson, Duc-Tien Dang-Nguyen, Jakub Lokoc, Luca Rossetto, Minh-Triet Tran, Wolfgang Hurst, Werner Bailer, Klaus Schoeffmann
Summary: This article presents the background and goals of the sixth Lifelog Search Challenge (LSC), including the participating teams and the organization of the competition.
PROCEEDINGS OF THE 2023 ACM INTERNATIONAL CONFERENCE ON MULTIMEDIA RETRIEVAL, ICMR 2023
(2023)