Find research datasets worth reusing

Source-Agnostic Gravitational-Wave Detection with Recurrent Autoencoders: H1 detector

Gravitational Wave signals and random noise datasets, generated with the PyCBC library (https://pycbc.org).  Data consists of 8 sec time sequences for a single detector (the H1 detector, mimicking LIGO Hanford), sampled at 2048 Hz and  represented as a one-dimensional array with 16,384 entries.  Signal samples are provided, corresponding to Binary Black Hole and Binary Neutron Star mergers overlapped to noise.

Source-Agnostic Gravitational-Wave Detection with Recurrent Autoencoders: L1 detector

Gravitational Wave signals and random noise datasets, generated with the PyCBC library (https://pycbc.org).  Data consists of 8 sec time sequences for a single detector (the L1 detector, mimicking LIGO Livingston), sampled at 2048 Hz and  represented as a one-dimensional array with 16,384 entries.  Signal samples are provided, corresponding to Binary Black Hole and Binary Neutron Star mergers overlapped to noise.

openother-openJul 2021View details →

MSVermet/Cell-Nuclei-Detection-And-Segmentation-CENet-UNet: Cell nuclei detection and segmentation

Fully automated nuclei detection and segmentation with CE-Net and U-Net on the PanNuke dataset. Repository includes code and dataset.

Radar-based mesocyclone detections in Switzerland from 2016-2020

The data is owned by MeteoSwiss and the Laboratoire de Télédétection Environnementale, EPFL. It is linked to the research article: Feldmann et al., 2021, A Characterisation of Alpine Mesocyclone Occurrence, submitted to Weather and Climate Dynamics. Whenever using the dataset, please include a reference to this article. This dataset contains thunderstorm track data of thunderstorms containing a radar-detected mesocyclone in Switzerland from 2016-2020. Tracked variables include time, position, size, terrain altitude, track direction, track distance, rotational velocity, vorticity and depth of rotation, amongst others. The README.txt file contains a comprehensive list of all included variables and their units. The data is provided as a .csv file. The methods used to obtain this data are described in the corresponding publication (Feldmann et al., 2021).

Codebook for the analysis of focus group interview data collected as part of the DETECT project.

This is the codebook created for the analysis of focus group interview data collected as part of the <a href="http://detectproject.eu/">DETECT project</a>. In particular, nine interviews were conducted with primary and secondary school teachers in Finland, Italy, Spain and the UK in order to explore their perceptions of critical digital literacies and how these are manifested in their practices. The interviews were organised and conducted by the researchers in the respective local HEIs between February-June 2020. A total of 7 focus-groups interviews took place with a total number of 39 participating teachers (7 from Finland, 6 from the UK, 9 from Spain and 17 from Italy). The interviews were conducted both face to face (schools in Spain) and online (schools in Finland, Italy and the UK) due to pandemic-related restrictions imposed shortly after the start of the data collection period. The collected data were read and segmented and all interview excerpts relating to the Critical Digital Literacies framework sub-dimensions were marked and chosen for the analysis. The total number of marked segments was 666 in all nine focus group interviews. To ensure the reliability of the analyses, the researchers translated and collected a representative sample of codings (n=117 out of 666) for each sub-dimensions from each school and these were examined by all researchers in several consensus meetings and the final criteria for each sub-category were constructed together based on those discussions. This dataset accompanies the intellectual outputs of the DETECT project, including amongst others this report: Gouseti, A., Bruni, I., Ilomäki, L., Lakkala, M., Mundy, D., Raffaghelli, J., Ranieri, M., Roffi, A., Romero, M. and Romeu, T. (2021) Schools’ perceptions and experiences of critical digital literacies across four European countries - DETECT Report 2. Accessed at: <a href="http://doi.org/10.5281/zenodo.5070394">http://doi.org/10.5281/zenodo.5070394</a> Acknowledgement This project was funded by Erasmus+, KA2. Project Reference: 2019-1-UK01-KA201-061508.

opencc-by-4.0Jul 2020View details →

Detecting Cross-Language Plagiarism using Open Knowledge Graphs

Corresponding authors: <a href="mailto:meuschke@uni-wuppertal.de?subject=Inquiry%20about%20CL-OSA%20dataset">Norman Meuschke</a>, <a href="mailto:ruas@uni-wuppertal.de?subject=Inquiry%20about%20CL-OSA%20dataset">Terry Ruas</a> Venue: 2nd Workshop on Extraction and Evaluation of Knowledge Entities from Scientific Documents (EEKE2021) at the ACM/IEEE Joint Conference on Digital Libraries 2021 (JCDL2021) ========================================================================== Source code: <a href="https://github.com/ag-gipp/cl-osa">https://github.com/ag-gipp/cl-osa</a>  ========================================================================== Dataset Details <a href="https://jipsti.jst.go.jp/aspec/">ASPEC</a>. The Asian Scientific Paper Excerpt Corpus comprises excepts of scientific papers in Japanese that have been manually translated to English and Chinese. We use both subsets of the ASPEC corpus.  <ul> <li>ASPEC-JC contains abstracts and paragraphs from the main text of research papers that were translated manually from Japanese to Chinese.</li> <li>ASPEC-JE contains abstracts of approx. two million research papers that were translated manually from Japanese to English.  </li> </ul> <a href="https://ec.europa.eu/jrc/en/language-technologies/jrc-acquis">JRC-Acquis</a>. The corpus consists of legislative texts in 22 languages, which the European Union's Joint Research Centre (JRC) selected from the cumulative body of EU laws (the so called Acquis communautaire). We sampled our test cases from the 10,000 document pairs in the English-French subset of the corpus. <a href="https://www.statmt.org/europarl/">Europarl</a>. The corpus contains transcripts of European Parliament proceedings in 21 European languages. We exclusively sampled test cases from the 9,443 document pairs in the English-French subset of the corpus. <a href="https://zenodo.org/record/3250095#.YOr25jqxVH4">PAN-PC-11</a>. The corpus contains instances of simulated monolingual and cross-language plagiarism that were used for evaluating plagiarism detection methods as part of the workshop series <a href="https://pan.webis.de/">Plagiarism Analysis, Authorship Identification, and Near-Duplicate Detection (PAN)</a>. Most of the 26,939 documents in the corpus were created by extracting text from openly available books. The documents are partially interspersed with instances of simulated plagiarism that were created and obfuscated automatically or by crowdsourced workers. We exclusively sampled test cases from the 2,921 Spanish-English aligned document pairs in the corpus, for which simulated plagiarism instances were either machine-generated or created manually by crowdsourced workers. <ul> </ul> ========================================================================== File Structure [corpus_documents] folder: Corpora of translation-aligned documents used in our experiments composed of:  <ul> <li>aspec: Japanese and English</li> <li>aspecx: Japanese and Chinese </li> <li>jrc: English and French                  </li> <li>europarl: English and French</li> <li>pan: English and Spanish</li> </ul> Each sub-corpus consists of 4,000 translation-aligned files (2,000 per language); the entire corpus has thus 20,000 files. Each set of translation-aligned documents was randomly selected from the original datasets (details in the paper). The Japanese files in aspec and aspecx do not necessarily overlap even though they are from the same dataset. [vectors_documents] folder: Average vector representation of the documents in the datasets from two pre-trained models: <ul> <li>Universal Sentence Encoder - Multilingual (USE-ML)</li> <li>ConceptNet Numberbatch</li> </ul>   Naming convention: <model>_<dataset>_<language>; <ul> <li>Example: cn_jrc_es: <ul> <li>model: ConceptNet Numberbatch</li> <li>corpus: JRC-Acquis</li> <li>language: Spanish</li> </ul> </li> </ul> <ul> <li>Labels: <ul> <li><model>: cn - ConceptNet Numberbatch um - USE-ML  </li> <li><dataset> aspec - ASPEC (Asian Scientific Paper Excerpt Corpus) - English and Japanese aspecx - ASPEC (Asian Scientific Paper Excerpt Corpus) - Japanese and Chinese jrc - JRC-Acquis europarl - Europarl pan - PAN-PC-11  </li> <li><language> en - English es - Spanish fr - French ja - Japanese zh - Chinese</li> </ul> </li> </ul>

opencc-by-4.0Aug 2021View details →

WtaGraph: Web Tracking and Advertising Detection using Graph Neural Networks

Dataset release for our IEEE Symposium on Security and Privacy 2022 paper entitled "WtaGraph: Web Tracking and Advertising Detection using Graph Neural Networks" These files are pretty self-explanatory: there are node/edge features of a specified (full graph or random 5K graph). To use these files, make sure to check out GitHub <a href="https://github.com/jun521ju/IEEE_SP_2022_WtaGraph">here</a>. Basically, you need put each of these files in a correct folder as detailed on GitHub.

dryad40/100

Supplement: Multicenter validated detection of focal cortical dysplasia using deep learning

Objective. To test the hypothesis that a multicenter-validated computer deep learning algorithm detects MRI-negative focal cortical dysplasia (FCD). Methods. We used clinically acquired 3D T1-weighted and 3D FLAIR MRI of 148 patients (median age, 23 years [range, 2-55]; 47% female) with histologically verified FCD at nine centers to train a deep convolutional neural network (CNN) classifier. Images were initially deemed as MRI-negative in 51% of cases, in whom intracranial EEG determined the focus. For risk stratification, the CNN incorporated Bayesian uncertainty estimation as a measure of confidence. To evaluate performance, detection maps were compared to expert FCD manual labels. We also tested sensitivity in an independent cohort of 23 FCD cases (13±10 years). Applying the algorithm to 38 healthy and 63 temporal lobe epilepsy disease controls tested specificity. Results. Overall sensitivity was 93% (137/148 FCD detected) using a leave-one-site-out cross-validation, with an average of six false positives per patient. Sensitivity in MRI-negative FCD was 85%. In 73% of patients, the FCD was among the clusters with the highest confidence; in half it ranked the highest. Sensitivity in the independent cohort was 83% (19/23; average of five false positives per patient). Specificity was 89% in healthy and disease controls. Conclusions. This first multicenter-validated deep learning detection algorithm yields the highest sensitivity to date in MRI-negative FCD. By pairing predictions with risk stratification this classifier may assist clinicians to adjust hypotheses relative to other tests, increasing diagnostic confidence. Moreover, generalizability across age and MRI hardware makes this approach ideal for pre-surgical evaluation of MRI-negative epilepsy. Classification of evidence. This study provides Class III evidence that deep learning on multimodal MRI accurately identifies FCD in epilepsy patients initially diagnosed as MRI-negative.

opencc-zeroAug 2021View details →

opencc-by-4.0Aug 2021View details →

Detecting Hogweed on the Ground-Level View Photographs: Dataset

Semantic segmentation highlighting Sosnowsky's hogweed on ground-level view images. Photographic images are hosted on Zenodo. COCO-formatted (*.json) polygons are to be published on Github: https://github.com/alexeyev/hogweed-ground-level-view

opencc-by-4.0Dec 2020View details →

Diagnostic accuracy of a novel enzyme‑linked immunoassay for the detection of IgG and IgG4 against Strongyloides stercoralis based on the recombinant antigens NIE/SsIR.

Background: The diagnosis of strongyloidiasis is challenging. Serological tests are acknowledged to have high sensitivity, but issues due to cross-reactions with other parasites, native parasite antigen supply and intrinsic test variability do occur. Assays based on recombinant antigens could represent an improvement. The aim of this study was to assess the sensitivity and specificity of two novel immunoglobulin (Ig)G and IgG4 enzyme-linked immunosorbent assays (ELISAs) based on the recombinant antigens NIE/SsIR for the diagnosis of strongyloidiasis. Methods: This was a retrospective diagnostic accuracy study. We included serum samples collected from immigrants from strongyloidiasis endemic areas for whom there was a matched result for Strongyloides stercoralis on agar plate culture and/or PCR assay, or a positive microscopy for S. stercoralis larvae. For the included samples, results were also available from an in-house indirect fluorescent antibody test (IFAT) and a commercial (Bordier ELISA; Bordier Affinity Products SA) ELISA. We excluded: (i) samples with insufficient serum volume; (ii) samples from patients treated with ivermectin in the previous 6 months; and (iii) sera from patients for whom only routine coproparasitology was performed after formol–ether concentration, if negative for S. stercoralis larvae. The performance of the novel assays was assessed against: (i) a primary reference standard, with samples classified as negative/positive on the basis of the results of fecal tests; (ii) a composite reference standard (CRS), which also considered patients to be positive who had concordant positive results for the IFAT and Bordier ELISA or with a single “high titer” positive result for the IFAT or Bordier ELISA. Samples with a single positive test, either for the IFAT or Bordier ELISA, at low titer, were considered to be “indeterminate,” and analyses were carried out with and without their inclusion. Results: When assessed against the primary reference standard, the sensitivities of the IgG and IgG4 ELISAs were 92% (95% confidence interval [CI]: 88–97%) and 81% (95% CI: 74–87%), respectively, and the specificities were 91% (95% CI: 88–95%) and 94% (95% CI: 91–97%), respectively. When tested against the CRS, the IgG ELISA performed best, with 78% sensitivity (95% CI: 72–83%) and 98% specificity (95% CI: 96–100%), when a cut-off of 0.675 was applied and the indeterminate samples were excluded from the analysis. Conclusion: The NIE-SsIR IgG ELISA demonstrated better accuracy than the IgG4 assay and was deemed promising particularly for serosurveys in endemic areas.

Fig. 5 in Sampling mammalian carnivores in western Thailand: Issues of rarity and detectability

Fig. 5. Model averaged estimates of detection probability for 15 carnivore species by camera trapping plotted as a function of log10(weight in kg). Error bars represent 95% confidence intervals. Fitted line is derived from a second-order polynomial which described the best fit of the data. R2 is the coefficient of determination of the fitted line. The six species with lg(weight) greater than 1.0 are Asiatic golden cat, Asiatic jackal, dhole, sun bear, Asiatic black bear, and tiger, from left to right.

Fig. 4 in Sampling mammalian carnivores in western Thailand: Issues of rarity and detectability

Fig. 4. Model averaged estimates of detection probability across the models shown in Tables 2B and 2C for species functional groups (TRAIT; large vs small and terrestrial vs semi-arboreal behavior). Estimated detection probabilities of each species are also shown for TERRESTRIAL and SEMI-ARBOREAL species. Error bars represent 95% confidence intervals.

Fig. 3 in Sampling mammalian carnivores in western Thailand: Issues of rarity and detectability

Fig. 3. Model averaged estimates of detection probability of four functional groups of carnivores (top row) and for each of 15 species separately. Detection probabilities are estimated as functions of camera placement categories (road [no. cameras = 1], streams [5], trails [14], & trails by streams [tbs; 33]). Error bars represent 95% confidence intervals. Estimates were presented without error bars when standard errors could not be estimated. Estimates with 95% confidence intervals less than zero and/or greater than one indicate a lack of model convergence. n refers to the number of independent photographs used to estimate detection probabilities.

Fig. 2 in Sampling mammalian carnivores in western Thailand: Issues of rarity and detectability

Fig. 2. Species discovery curves generated from rarefaction for mammalian carnivores based on camera trap surveys at four study sites within mosaic forest types of Thung Yai between November 2007 and August 2008. Bars represent 95% confidence intervals of the estimates scaled by survey effort (camera trap nights) combined across all active camera locations (n = 53). Curves were estimated using package BiodiversityR in program R.

Fig. 1 in Sampling mammalian carnivores in western Thailand: Issues of rarity and detectability

Fig. 1. Map of Thung Yai Naresuan Wildlife Sanctuary showing camera trap polygons in four study sites, from south to north, Headquarters (HQ), Tikong (TK), Sesawo (SSW), and Mae Gatha (MGT) between November 2007 and August 2008.

opencc-by-4.0Sep 2021View details →

Super Code Clone Detection - 88 (SCD-88)

SCD-88 is the Python-specific subset of the <a href="https://ieeexplore.ieee.org/document/8816761/">Cross-Language Clone Detection dataset</a> which was originally extracted from <a href="https://atcoder.jp/">AtCoder</a>, a popular Online Judge. We reformulate this classification task as a retrieval one where given a code and a collection of candidates as the input, the task is to return top-k codes with the same semantic. Models can hence, be evaluated by the MAP@R score. MAP@R is defined as the mean of average precision scores, each of which is evaluated for retrieving R most similar samples given a query. For a code (query), R is the number of other codes in the same class, i.e. R=129 in this dataset. The newly sampled dataset amounts to a total of 11,440 examples where the splits are as follows: 7800 / 1040 / 2600 (Train / Valid / Test).

opencc-by-4.0Sep 2021View details →

Public resources about PHYVV and TMV detection in Jalapeño pepper plants using CNNs

This repository contains a dataset of Jalapeño pepper leaves which where divided into three classes labeles as "healthy", "PHYVV-infected" and "TMV-infected. In addition, the folder labeled as model contains all the information of the artificial intelligence model that was utilized to classify between the aforementioned classes.

opencc-by-4.0Sep 2021View details →

Data from: Detecting the effects of rapid tectonically-induced subsidence on Mayotte Island since 2018 on beach and reef morphology, and implications for coastal vulnerability to marine flooding

This dataset contains data from the monitoring morphological evolution of beaches and coral reefs in Mayotte island.  Mayotte, part of the coral reef-fringed Comoro archipelago in the SW Indian Ocean, experienced in 2018 and 2019 an intense seismic crisis. The repeated earthquake activity since May 2018 has been associated with deformation of the surface of Mayotte, resulting in land subsidence. The earlier 2006-2008 profiles were realized using a Leica TC 407® total station, and referenced to local IGN 50 benchmarks. The more recent 2019, 2020, and 2021 surveys were carried out using a GNSS differential Trimble R8S® system. Given the rapid subsidence that has affected Mayotte, the benchmarks used in this study, like others in Mayotte, need to be recalibrated by the IGN (French Institut Géographique National) and SHOM. This has still not yet been done, as the final outcome of the vertical island movements is still not clear.

opencc-by-4.0Aug 2020View details →

Series AC Arc Fault Detection Method Based on High-Frequency Coupling Sensor and Convolution Neural Network

The data provided can be used for the development of methods for the detection of arcing faults in a domestic low-voltage electrical networks (230V - 50 Hz). The data files are current and voltage signatures experimentally measured. Test for to produce an arcing fault : Open contact electrodes and Carbonized path wires The ReadMe file describes : - the test set up and the  the procedure followed to make the measurements - the list of household appliances and their main characteristics. - the name of the data files - the type of arcing faults