Find research datasets worth reusing
Search datasets from major research repositories and use ShareScore to quickly assess how well each record supports discovery, access, and reuse.
101
datasets available to search
ShareScore release 0.9.0
Dataset results
101 results for “workload”
Edge workload traces from Aeneas, Julius, and MR-Leo
<p><strong>Overview</strong></p> <p>The emerging field of edge computing is suffering from a lack of representative data to evaluate rapidly emerging new algorithms or techniques. It is a critical issue as this complex paradigm has numerous different use cases which translate to a highly diverse set of workload types.</p> <p>In this dataset, we collect traces for three available edge applications: Aeneas [1], Julius [2], and MR-Leo [3]. They are provided open-source. Read our article [4] for explanations about the gathering of these traces and an analysis of them.</p> <p>[1] https://www.readbeyond.it/aeneas/<br> [2] https://github.com/julius-speech/julius<br> [3] https://gitlab.liu.se/ida-rtslab/public-code/2019_mrleo_server<br> [4] K. Toczé, N. Schmitt, U. Kargén, A. Aral, and I. Brandic, <strong>Edge Workload Trace Gathering and Analysis for Benchmarking</strong>, in <em>6th IEEE International Conference on Fog and Edge Computing 2022 (ICFEC 2022)</em>, IEEE, 2022. DOI: <a href="https://doi.org/10.1109/ICFEC54809.2022.00012">10.1109/ICFEC54809.2022.00012</a></p> <p><strong>Details about the hardware used for the trace collection</strong></p> <p>The Aeneas trace was gathered using a an HP Elitebook 840 G5 running Ubuntu 18.10. It has 16 GB RAM and an Intel Core i7-8550U CPU (1.8 GHz, 4 cores, 8 threads). Aeneas version 1.7.3 was used.</p> <p>We collect traces for Julius on four Off-The-Shelve servers with varying performance as depicted in the following table. Because Julius is a single-core application, performance is mostly dictated by the clock speed of the CPU. Each server is running Ubuntu 18.04.4 LTS with kernel version 4.15.0-108-generic.<br> As some audio files caused Julius to crash, all files are converted from the original Ogg Vorbis format to WAV by applying \texttt{ffmpeg -i audio.ogg -acodec pcm\_s16le -ac 1 -ar 16000}, resulting in a total of 5724 audio files.<br> </p> <table> <caption>System under test servers for the Julius speech recognition application. (All CPUs are Intel Xeon)</caption> <tbody> <tr> <td>Server</td> <td>CPU @ Clock (Cores / Threads)</td> <td>Memory @ Clock</td> </tr> <tr> <td>A</td> <td>E3-1230 v5 @ 3.4GHz (4/8)</td> <td>1x16GB @ 2133MHz</td> </tr> <tr> <td>B</td> <td>E5-2640 v3 @ 2.6GHz (8/16)</td> <td>2x16GB @ 2133MHz</td> </tr> <tr> <td>C</td> <td>E5-2650 v3 @ 2.3GHz (10/20)</td> <td>2x16GB @ 2133MHz</td> </tr> <tr> <td>D</td> <td>E5-2650 v4 @ 2.2GHz (12/24)</td> <td>2x16GB @ 2133MHz</td> </tr> </tbody> </table> <p>The MR-Leo trace is gathered using a an HP Elitebook 840 G5 running Ubuntu 18.10. It has 16 GB RAM and an Intel Core i7-8550U CPU (1.8 GHz, 4 cores, 8 threads). The MR-Leo implementation using ORB-SLAM2 is used, in the replay mode on the edge server, meaning that the input video was directly streamed on the edge server, and not from an end device.</p> <p><strong>Details about the task id field</strong></p> <p>The task id for the Aeneas trace looks as follows: A_1_3 where the A stands for Aeneas, the first number corresponds to the example used for this task (see Figure 3 in parenthesis) and the second number identify the run number in which the data is collected.</p> <p>The task id field in the trace file uses the format J_C_3_4 where J stands for Julius, the second letter corresponds to the server on which the measurement has been taken (see Table above), the first number is the file that is converted (see Figure 4), and the second number is the run number.</p> <p>The task id field of the MR-Leo trace looks as follows: M_1_4_78. The M stands for MR-Leo, the first number identifies the video, the second number identifies the run number and the last number is the frame number.</p>
WESSBAS: Extraction of Probabilistic Workload Specifications for Load Testing and Performance Prediction - A Model-Driven Approach for Session-Based Application Systems.
<p>Supplementary material for the paper: "WESSBAS: Extraction of Probabilistic Workload Specifications for Load Testing and Performance Prediction".</p> <p>Included in the supplementary material are the evaluation results.</p> <p>The WESSBAS software relevant to the paper is available via https://github.com/Wessbas/</p> <p>The WESSBAS UI is available as a password-protected (password: wessbasui) ZIP file:</p> <p>https://dl.dropboxusercontent.com/u/81621779/wessbas.ui.zip (--- WESSBAS GUI (license confirmation pending, i.e., not on GitHub, yet))</p>
The impact of deep learning aid on the workload and interpretation accuracy of radiologists on chest computed tomography: a cross-over reader study.
<p>Data used to perform statistical analysis in "The impact of deep learning aid on the workload and interpretation accuracy of radiologists on chest computed tomography: a cross-over reader study.". </p>
Dataset for Characterizing Distributed Machine Learning Workloads on Apache Spark
<div> <div> <div> <ul> <li> <p><span>YasmineDjebrouni,IsabellyRocha,SaraBouchenak,LydiaChen,PascalFelber,Vania Marangozova, and Valerio Schiavoni. 2023. Characterizing Distributed Machine Learning Workloads on Apache Spark. In Proceedings of the 24th International Middleware Conference (Middleware ’23). Association for Computing Machinery, New York, NY, USA, 151–164. </span></p> </li> </ul> </div> </div> </div>
Energy consumption, execution time and fail requests rate of a proactive energy-aware auto-scaling solution for edge-based infrastructures applied to real-world workload.
<p>Spreadsheet of the results obtained with our horizontal auto-scaling proposal presented in "A proactive energy-aware auto-scaling solution for edge-based infrastructures". In that research, we present a proactive horizontal auto-scaling framework for edge infrastructures, which considers both the base (idle) and dynamic (due to application execution) energy consumption of edge nodes and the node scaling mechanism. Simulations were performed with the EdgeCloudSim simulator with a workload provided by Shanghai Telecom and the results show up to a 92.5% decrease in energy consumption, a failed request rate of up to 0%, and reasonable execution times of the auto-scaling process for different problem sizes.</p> <p>Proactive auto-scaling mechanisms in edge-based infrastructures can anticipate user service requests by allocating computing resources while supporting the quality of service needed by a vast range of applications requiring, e.g., a low latency or response time. </p> <p>This work is supported by the European Union's H2020 research and innovation program under grant agreement DAEMON 101017109 and by the projects co-financed by FEDER funds LEIA UMA18-FEDERJA-15, MEDEA RTI2018-099213-B-I00 (MCI/AEI) and RHEA P18-FR-1081.</p>
The workload of manual data entry for integration between mobile health applications and eHealth infrastructure
<div> <p>In this study, we conducted a time-motion study observing healthcare workers (HCWs) completing data management activities including monitoring and evaluation (M&E) and manual data linkage of individual-level app data to electronic medical records (EMRS). This study served as a baseline study for an open-source app to mirror EMRS and reduce HCW workload while improving care in the Nurse-led Community-based Antiretroviral therapy Program (NCAP) in Lilongwe, Malawi.</p> </div>
Artificial Software Testing Workload
<p>This is a small dataset representing an artificial software testing workload generated from the real one. It consists of instance files, each of which is structured as follows:</p> <pre><code>[number of test suites] [test suite uuid] [test suite priority (1>0)] [arrival time (s)] [number of test cases] [test case uuid] [test case type] [test case outcome (0 - pass)] [test case duration (s)] [test case uuid] [test case type] [test case outcome (0 - pass)] [test case duration (s)] (...) [test suite uuid] [test suite priority (1>0)] [arrival time (s)] [number of test cases] [test case uuid] [test case type] [test case outcome (0 - pass)] [test case duration (s)] [test case uuid] [test case type] [test case outcome (0 - pass)] [test case duration (s)] (...) (...)</code></pre>
Dataset for nonlinear dynamic analysis to quantify workload in pedaling
<p>The dataset contains raw data and the data underlying the figures and tables of the following research article: </p> <p>"Quantifying workload using nonlinear dynamical measures of biomechanical parameters during cycling on a roller trainer"</p> <p>Journal: PLOS ONE</p> <p>Manuscript Nr: PONE-D-22-23944</p>
Virtual Forest Twins based on Marteloscope point cloud data from different forest ecosystem types and its associated teaching workload
<p>The 8th intellectual output (IO-8) titled "Virtual Forest Twins based on Marteloscope point cloud data from different forest ecosystems type and its associated teaching workload" [of the Erasmus+ project "Virtualization of Forest Studies (Virtual Forests)"] was led by Eberswalde University for Sustainable Development (HNEE) in Eberswalde, Germany. The main objectives of IO-8 were two-fold: (1) the development of an analytical pipeline for the virtualization of real forest stands from selected marteloscope sites into 3D virtual twins, and (2) the preparation of open-access, blended teaching and training materials based on the aforementioned analytical pipeline. Specifically, 3D virtual forest twins were created based on point cloud data derived from UAV-mounted and hand-held terrestrial laser scanning technology, followed by a series of software- and hardware-supported analytical steps, and finally rendered for visualization and presentation. The point cloud data collected represents the first significant product of IO8. All further results and products are compiled into open-source resources for teaching, training and research purposes in higher education and forest science. These include technical guidelines, MSc.-level learning materials available for self-directed study on the project Moodle, and a stand-alone online multimedia tutorial narrated in a published ArcGIS StoryMap. </p>
Neurological Complications and ICU Workload After Emergency Repair of Acute Type A Aortic Dissection
ClinicalTrials.gov study NCT07167628. IPD Sharing: NO. Countries: 1. Publications: 10.
The workload of manual data entry for integration between mobile health applications and eHealth infrastructure
Open the record for dataset details and reuse information.
Perceived Mental Workload Classification using Intermediate Fusion Multimodal Deep Learning
<p><em>This repository contains all code -from data collection to perceived mental workload classification- used in the "Perceived Mental Workload Classification using Intermediate Fusion Multimodal Deep Learning" research and serves as supplementary material. <strong>A clear README file is provided, please refer to that for information about the individual scripts and their usage. </strong></em></p> <p>Mental workload detection has been attempted using various bio-signals. Recently, deep learning has allowed for novel methods and results within the BCI community. However, studies currently often only use a single modality to classify mental workload, whereas a plethora of modalities have proven to be valuable in this task.</p> <p>A dataset on which these scripts was also made publicly available under the following DOI: 10.4121/12932801<br> This dataset contains data collected during research into mental workload (MWL) detection using deep learning. It is being made public as supplementary data for publications, as well as for reuse in research that seeks to classify MWL using multimodal physiological data. The goal of this repository and dataset is to serve as a testing ground for the creation of deep neural networks that can classify MWL using multimodal physiological data.</p> <p>The data in this dataset was collected in the Behavioural, Management, and Social Sciences Lab, University of Twente, Enschede, The Netherlands in June/July 2020.</p>
Supplementary Material to Context-tailored Workload Model Generation for Continuous Representative Load Testing
<p>This package contains files and instructions to replicate our experiments on context-tailored workload model generation for continuous representative load testing. Please download and extract the <code>context-tailoring.tar.gz</code> file and view <code>index.html</code> (also contained in the tarball) for further instructions.</p> <p><code>cobra-db.tar.gz </code>is only required for skipping parts of the steps (see the detailed instructions).</p>
Data from: Individual variation in parental workload and breeding productivity in female European starlings: is the effort worth it?
We analyzed individual variation in work load (nest visit rate) during chick-rearing, and the consequences of this variation in terms of breeding productivity, in a highly synchronous breeder, the European starling (Sturnus vulgaris) focusing on female birds. There was marked (10- to 16-fold) variation in total, female and male nest visit rates, among individuals, but individual variation in female nest visit rate was independent of environment (rainfall, temperature) and metrics of individual quality (laying date, clutch size, amount of male provisioning help), and was only weakly associated with chick demand (i.e., day 6 brood size). Female nest visit rate was independent of date and experimentally delayed birds provisioned at the same rate as peak-nesting birds; supporting a lack of effect of date per se. Brood size at fledging was positively but weakly related to total nest visit rate (male + female), with >fivefold variation in nest visit rate for any given brood size, and in females brood size at fledging and chick mass at fledging were independent of female nest visit rate, that is, individual variation in workload was not associated with higher productivity. Nevertheless, nest visit rate in females was repeatable among consecutive days (6–8 posthatching), and between peak (first) and second broods, but not among years. Our data suggest that individual females behave as if committed to a certain level of parental care at the outset of their annual breeding attempt, but this varies among years, that is, behavior is not fixed throughout an individual's life but represents an annually variable decision. We suggest females are making predictable decisions about their workload during provisioning that maximizes their overall fitness based on an integration of information on their current environment (although these cues currently remain unidentified).
Characterizing Distributed Machine Learning Workloads on Apache Spark
<p>This dataset was used for our submission at Middleware'22 titled: "Characterizing Distributed ML Workloads"</p> <p>It will contains the description and the raw data, its format, as well as a detailed description of the cluster deployments used by these experiments.<br> </p> <p>The full paper is available here:</p> <p>https://dl.acm.org/doi/10.1145/3590140.3629112</p>
UNIVERSE: UNobtrusIVE measuRement of mental workload and stress in uncontrolled environments
<p><strong>A Dataset on Unobtrusive Measurement of Cognitive Load and Physiological Signals (EEG, PPG, EDA) in Uncontrolled Environments</strong></p> <p>The dataset (approximately 315 hours in total) consists of physiological signals from wearable electroencephalography (EEG), electrodermal activity (EDA), photoplethysmogram (PPG), acceleration, and temperature sensors. The recorded dataset is curated from 24 participants following an eight-hour cognitive load elicitation paradigm. The mentioned consumer-grade physiological signals are obtained from the Muse S EEG headband and Empatica E4 wristband. The data is balanced across controlled and uncontrolled environments and high vs. low mental workload levels. During the study, participants worked on mental arithmetic, Stroop, N-Back, and Sudoku tasks in the controlled environment (roughly half of the data) and realistic home-office tasks such as researching, programming, and writing emails in uncontrolled environments. Data labels were obtained using Likert scales, Affective Sliders, PANAS, and NASA-TLX questionnaires. The completely anonymized data set and its publicly available features open a vast potential to the research community working on mental workload detection using consumer-grade wearable sensors. Among others, the data is suitable for developing real-time cognitive load detection methods, research on signal processing techniques for challenging environments, developing artifact removal techniques from low-cost wearable devices' data, or developing personal mental workload assistants.<br> </p> <p>Follow the dataset descriptor publication for more details:</p> <p>Anders, C., Moontaha, S., Real, S., Arnrich, B.: Unobtrusive measurement of cognitive load and physiological signals in uncontrolled environments. Scientific Data 11(1), 1000 (2024).</p> <p><a href="https://doi.org/https://doi.org/10.1038/s41597-024-03738-7" target="_blank" rel="noopener">https://doi.org/https://doi.org/10.1038/s41597-024-03738-7</a></p>
Raw Experimental Data of the Master's Thesis: Improving Serverless Edge Computing for Network Bound Workloads
<p>Raw Experimental Data of the Master's Thesis: Improving Serverless Edge Computing for Network Bound Workloads</p> <p>Please see the README for information on parsing and data structure.<br> Some of the data might need additional explanations. In case of confusion don't hesitate to contact me under jacob.palecek@outlook.com</p>
Energy Efficient Portfolio Scheduling for Business-Critical Workloads
<p>Results and data needed for reproducing the experiments run in https://github.com/Timovanmilligen/opendc defined in the folder 'opendc-experiments-timo'</p>
F-DATA: A Fugaku Workload Dataset for Job-centric Predictive Modelling in HPC Systems
<p>F-DATA is a novel workload dataset containing the data of around 24 million jobs executed on <a href="https://www.r-ccs.riken.jp/en/fugaku/">Supercomputer Fugaku</a>, over the three years of public system usage (March 2021-April 2024). Each job data contains an extensive set of features, such as exit code, duration, power consumption and performance metrics (e.g. #flops, memory bandwidth, operational intensity and memory/compute bound label), which allows for a multitude of job characteristics prediction. The full list of features can be found in the file <code>feature_list.csv</code>.</p> <p>The sensitive data appears both in anonymized and encoded versions. The encoding is based on a Natural Language Processing model and retains sensitive but useful job information for prediction purposes, without violating data privacy. The scripts used to generate the dataset are available in the<a href="https://github.com/francescoantici/F-DATA"> F-DATA GitHub repository</a>, along with a series of plots and instruction on how to load the data.</p> <p>F-DATA is composed of 38 files, with each <code>YY_MM.parquet</code> file containing the data of the jobs submitted in the month MM of the year YY. </p> <div> <div>The files of F-DATA are saved as <code>.parquet</code> files. It is possible to load such files as dataframes by leveraging the <code>pandas</code> APIs, after installing <code>pyarrow</code> (<code>pip install pyarrow</code>). A single file can be read with the following <code>Python</code> instrcutions:</div> <br> <blockquote> <div><code># Importing pandas library</code></div> <div><code>import pandas as pd</code></div> <div> </div> <div><code># Read the 21_01.parquet file in a dataframe format</code></div> <div><code>df = pd.read_parquet("21_01.parquet")</code></div> <div><code>df.head()</code></div> </blockquote> <div> </div> <div>Please cite this work as:<br><br> <div> <div>@article{antici2025fdata,</div> <div>title={F-DATA: A Fugaku Workload Dataset for Job-centric Predictive Modelling in HPC Systems},</div> <div>author={Antici, Francesco and Bartolini, Andrea and Domke, Jens and Kiziltan, Zeynep and Yamamoto, Keiji},</div> <div>journal = {Scientific Data},</div> <div>volume={12},</div> <div>pages={1321},</div> <div>year={2025},</div> <div>publisher={Nature Publishing Group},</div> <div>doi={https://doi.org/10.1038/s41597-025-05633-1}</div> <div>}</div> </div> </div> </div>
Replication Package: Microservice-tailored Generation of Session-based Workload Models for Representative Load Testing
<p>This is the replication package for the publication <em>Microservice-tailored Generation of Session-based Workload Models for Representative Load Testing</em>, MASCOTS 2019. It holds the experiment setup, results, and detailed analyses of the results.</p> <p>The README.md (or README.pdf) contains further descriptions and instructions.</p>
ScienceDex guides
Understand access before you commit
These curated guides explain access requirements, typical timelines, costs, and reuse considerations for widely used research datasets.
Allen Brain Atlas
Allen Brain Atlas is an Allen Institute collection of brain map atlases, datasets, APIs, and analysis tools covering mouse, human, and non-human primate brain resources.
Annotated Behaviour and Observability Dataset (ABODe)
ABODe is a University of Edinburgh DataShare dataset for behavior classification in group-housed mice using home-cage video, identities, bounding boxes, ground-plate positions, and annotator labels.
DANDI Archive for NWB datasets
DANDI is a BRAIN Initiative archive for publishing and sharing neurophysiology data, including electrophysiology, optophysiology, and behavioral data packaged as NWB and related standards.
International Brain Laboratory public data
The International Brain Laboratory public data releases expose standardized mouse decision-making experiments, including Neuropixels recordings, widefield calcium imaging, behavior, and session metadata accessed through the ONE API.
OpenNeuro
OpenNeuro is a free, open platform for sharing neuroimaging datasets, with public search, dataset pages, and download paths for web, S3, DataLad, and the OpenNeuro CLI.