Find research datasets worth reusing

opencc-by-4.0Jul 2023View details →

Maintenance of Plan Libraries for Case-based Planning

Case-based planning is an approach to planning where previous planning experience provides guidance to solving new problems. Such a guidance can be extremely useful, or even necessary, when the new problem is very hard to solve, or the stored previous experience is highly valuable, because, e.g., it was provided or validated by human experts, and the system should try to reuse it as much as possible.  To do so, a case-based planning system stores in a library previous planning experience in the form of already encountered problems and their solutions.  The quality of such a plan library critically influences the performance of the planner, and therefore it needs to be carefully designed and created. For this reason, it is also important to update the library during the lifetime of the system, as the type of problems being addressed may evolve or differ from the ones the library was originally designed for. Moreover, like in general case-based reasoning, the library needs to be maintained at a manageable size, otherwise the computational cost of querying it grows excessively, making the entire approach ineffective.

opengpl-2.0Aug 2023View details →

SPEChpc 2021 Benchmarks: A Performance and Energy Case Study

SPEChpc 2021 Benchmarks on Ice Lake and Sapphire Rapids based Infiniband Clusters: A Performance and Energy Case Study.

opencc-by-4.0Sep 2023View details →

Report on Transformers interpretability for Natural Language Processing: A case study on Technical Debt classification

Transformer models have significantly advanced the field of natural language processing (NLP), achieving exceptional results in various tasks. However, these models are often seen as "black boxes", providing limited insight into the factors influencing their predictions. It has become crucial to develop and utilise methods for interpreting and explaining these models to uncover their complex inner workings. This report discusses the latest techniques and tools that aid in a more profound understanding of transformer models within NLP. Additionally, it explores a vital industrial use case: Technical Debt (TD) classification. In this context, the report leverages transformer model interpretability tools and Retrieval Augmented Generation (RAG) to analyse and understand the characteristics of text in Github issues, distinguishing between TD and non-TD. This report thoroughly outlines an approach to improve the transparency and reproducibility of machine learning models, with a special emphasis on TD classification. It integrates the RAG approach and exploits feature attribution techniques, presenting a route to create AI systems that are not only high-performing but also demonstrably trustworthy and comprehensible. Through a detailed examination of word patterns in TD classification and the innovative use of the RAG approach, the research highlights a strong dedication to promoting transparency and responsibility in AI systems, potentially ushering in a new phase in machine learning research that focuses on clarity and dependability.

opencc-by-4.0Sep 2023View details →

Film Programming in the USSR: A Case Study of Moscow Cinemas (1946–1955)

This paper presents a database on film programming in Moscow cinemas between 1946 and 1955. It outlines the place of this research at the intersection of new cinema history and academic debates on film distribution in the field of Soviet history. The paper describes the data collection, the coding to present the data, and the structure of the database, which consists of the three datasets on Moscow film programming (1946–1955), Moscow cinemas (1946–1955), and the 1952 film calendar. Concluding remarks summarize the knowledge obtained from the database and introduce the Soviet case into the international context of digital data collections for historical cinema studies.

opencc-by-4.0Sep 2023View details →

H2020 Platone German Demonstrator - Use Case Setting Data

This dataset belongs to the German demonstrator of the H2020 Platone project (WP5). The dataset contain information that have been set for Use Cases (UCs) parameterization during the project phase. UC have been parameterized along a grafical user interface (GUI) named "Use Case Selector". Each parameterized UC triggered has been logged in the dataset. Background - Field Test Setup The field test setup consists of a Low Voltage (LV) community with 450 kW installed generation capacity. The power exchange between the LV grid and Medium Voltage (MV) grid takes place along a single point of common coupling (PCC). i.e., a secondary substation that includes a transformer with senors on the LV busbar, to measure the net power exchange. The community consists of 89 households, 450kW of installed PV generation capacity, a Community Battery Energy Storage (CBES) connected to the LV busbar with 300 kW and 850 kWh capacity.  Definition of data: RequID - Request ID - Identifier for each triggered UC Alert - Indicates, whether UC has been executed successfull ( " "and " true" indacates successfull implementation by Energy Management System (EMS); "false" indicates that UC has not been implemented by EMS) Submission - timestamp of UC submission  Note - Annotations entered by UseCase operator Priority - Defines UC priority set by operator (priority: 1 - high , 2 - medium, 3 - low, 4 - very low) (only relevant for UC 2) Status - Indicates the status of the UC (closed - UC has been executed, cancelled, UC has been has been canceled before or during application) Start - Point of time set for the beginning of UC End - Point of time set for the end of UC Type- Triggered Type of UC (1 - "Virtual Islanding of LV community" (UC 1); 2 - "Coordination of Flex Request" (UC 2); 3 - "Energy Import in Bulk" (UC 3); 4 - "Bulk-based Energy Export" (UC 4) Subtype - 0 - Rule-Based Operation Mode with 15-minutes control cycles of battery (CBES in the field) ;1 - Day-ahead forecast-based control; 2.0 - Schedule-based operation mode with optimization applied to a day-ahead forecast (optimization target: minimization of power exchanges at MV/LV PCC within 24h period ; 21 - Schedule-based operation mode with optimization applied to a day-ahead forecast (optimization target: minimization of power exchanges at MV/LV PCC and achieving a requested State of Charge (of CBES) at the end of UC_End; bulkStart - Point of time of start of energy bulk import or export (only relevant for UC 3 and 4) bulkEnd - Point of time of end of energy bulk import or export (only relevant for UC 3 and 4) bulk Energy - Amount of energy triggered to be imported or exported as bulk (only relevant for UC 3 and 4) Final_SOF - State of Charge (SOC) of CBES that should be achieved at end of UC (End) FlexDemand - Requested power exchange that should be achieved at MV/LV PCC (Only relevant for UC 2) Ptcb - Measured CBES charging power at poin of time of UC submission  Ptei - Measured power exchange at PCC at point of time of UC submission  SoC - State of Charge of CBES at point of time of UC submission  SoE - State of Energyof CBES at point of time of UC submission  ActiveSet - State of Energyof CBES at point of time of UC submission  maxSoC - Maximum SoC set for CBES at point of time of UC submission  minSoc - MinimumSoC set for CBES at point of time of UC submission  This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 864300.

opencc-by-4.0May 2023View details →

Enhanced Westermo dataset - Transformed and Modified for Test case Selection and Priorotization in the context of Continuous Integration and Reinforcement Learning.

Overview This repository contains a modified version of the existing, recently published dataset, Westermo. The initial dataset was gathered at Westermo Network Technologies AB, located in Västerås, Sweden. It encompasses over 1 Million verdicts obtained from testing embedded systems, collected over a span of more than 500 consecutive days of nightly testing. The dataset has been transformed and tailored specifically to cater to the research community, particularly for addressing challenges such as regression test selection, identification of flaky tests, and visualization of test results. The original dataset can be accessed through the reference provided in [1]. The Westermo dataset offers valuable historical information regarding the execution of test cases and their corresponding results. It serves as a valuable resource for evaluating and comparing different Test case Selection and Prioritization (TSP) techniques, enabling researchers to identify test cases that are more likely to fail during subsequent executions. Test cases in the dataset are characterized by attributes such as execution duration, previous last execution time, and the results of their recent executions. This dataset offers valuable historical information regarding the execution of test cases and their corresponding results. It serves as a valuable resource for evaluating and comparing different test case prioritization and selection techniques, enabling researchers to identify test cases that are more likely to fail during subsequent executions. Test cases in the dataset are characterized by attributes such as execution duration, previous last execution time, and the results of their recent executions. <table align="left"> <caption>Table 1: Dataset Overview</caption> <tbody> <tr> <td>Test Cases</td> <td>1855</td> </tr> <tr> <td>CI Cycles</td> <td>15,197</td> </tr> <tr> <td>Verdict</td> <td>1,036,818</td> </tr> <tr> <td>Failed</td> <td>5.03%</td> </tr> </tbody> </table>               However, the diversity and multitude of the features in the dataset can be irrelevant to some TSP approaches. This led us to perform a dataset conversion, where we customized Westermo to have the same features from Paint Control and IOF/ROL, two widely used datasets in Reinforcement Learning based TSP approaches. This conversion required the combination of multiple variables and generating the target ones. When it comes to generating the “LastResults” and “Cycle” values, further analysis was required and the data handling needed an in-depth understanding of how the nightly testing was conducted. This led us to investigate what a CI cycle is in their context, and we followed their definition of a session, stating that “a session is when we run a suite of tests on one test system with a certain software version and testware version”. When splitting the data according to the 9 different systems used, we were able to generate 9 different sub-sets that fit the CI context.   File Format The compressed .zip file contains 9 files, each one corresponding to each of the 9 systems. The datasets are available in CSV format, with the semicolon (;) serving as the delimiter. The columns included are represented in the table below along with their descriptions. <table> <caption>Table 2: Parameters of the dataset</caption> <thead> <tr> <th scope="col">Column Name</th> <th scope="col">Content</th> </tr> </thead> <tbody> <tr> <td>Id</td> <td>Unique numeric identifier of the test execution </td> </tr> <tr> <td>Name</td> <td>Unique numeric identifier of the test case</td> </tr> <tr> <td>Duration</td> <td>Approximated runtime of the test case</td> </tr> <tr> <td>CalcPrio</td> <td>Priority of the test case, calculated by the prioritization algorithm (output column, initially 0)</td> </tr> <tr> <td>LastRun</td> <td>Previous last execution of the test case as date-time-string (Format: YYYY-MM-DD HH:ii )</td> </tr> <tr> <td>LastResults</td> <td>List of previous test results (Failed: 1, Passed: 0), ordered by ascending age. Lists are delimited by [ ].</td> </tr> <tr> <td>Verdict</td> <td> Test verdict of this test execution (Failed: 1, Passed: 0) </td> </tr> <tr> <td>Cycle</td> <td>The number of the CI cycle this test execution belongs to.</td> </tr> </tbody> </table>   The implications of this conversion are important as it can help the previous works to re-assess their approaches and have more data for training and testing, as well as opening a broader data spectrum for future researchers in this field to find ready-to-use, rich datasets, on which they could evaluate their approaches and contribute to the TSP community. This also addresses the limitations in the field discussed in the systematic literature review [2], stating that future research on TSP techniques should focus on collecting data from more recent subjects in a CI context with varying failure rates and larger execution times, as reproducible studies with appropriate datasets are needed to develop a usable body of knowledge regarding TSP over time. We believe that this conversion of the Westermo dataset is our contribution to alleviating the gap for the RL-based approaches. The original dataset can be found <a href="https://sites.mdu.se/aidoart/results/open-source/test-results-dataset-westermo">here.</a>

H2020 Platone Italian Demonstrator Use Case 1-2 Market 3rd quarter 2022

areti_market_flexibility_TSO_requestes areti_market_flexibility_DSO_requestes areti_market_flexibility_Aggregator_bids areti_market_flexibility_settlement areti_market_flexibility_outcomes - TSO flexibility requests: <ul> <li>Starting Time</li> <li>Duration</li> <li>Market Type</li> <li>Market Session</li> <li>Flexibility Service Type</li> <li>Volumes</li> <li>Grid Area</li> </ul> - DSO flexibility requests: <ul> <li>Starting Time</li> <li>Duration</li> <li>Market Type</li> <li>Market Session</li> <li>Flexibility Service Type</li> <li>Volumes, Grid Area</li> </ul> - Aggregator bids: <ul> <li>Starting Time</li> <li>Duration</li> <li>Market Type</li> <li>Market Session</li> <li>Flexibility Service Type</li> <li>Volumes</li> <li>PoDs List</li> </ul> - Settlement data: <ul> <li>Pod</li> <li>Requested Active Power</li> <li>Measured Active Power</li> <li>Requested Reactive Power</li> <li>Measured Reactive Power</li> </ul> - Market Outcomes: <ul> <li>Market Outcome Id</li> <li>Market Type</li> <li>Market Session</li> <li>Flexibility Service Type</li> </ul> Other than TSO flexibility requests, to test the demo, other data could be simulated. In this case, it will be indicated in the metadata documentation. (Useful link to consult Italian UC: <a href="https://eur01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fsmart-grid-use-cases.github.io%2Fdocs%2Fusecases%2Fplatone%2Fuc-it-1-voltage-management%2F&data=04%7C01%7Cfabio.bastianelli%40mail-bip.com%7C0bb5a42aac0c49cb876708d9f2c89510%7Cbb1a63ebeb09471aa00537b07792a5b5%7C0%7C0%7C637807765572787908%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&sdata=ptAsS52VBberHmZqIzYEXZs1PQrXQ6TDz6mNK%2FNWnk0%3D&reserved=0">https://smart-grid-use-cases.github.io/docs/usecases/platone/uc-it-1-voltage-management/</a>; <a href="https://eur01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fsmart-grid-use-cases.github.io%2Fdocs%2Fusecases%2Fplatone%2Fuc-it-2-congestion-management%2F&data=04%7C01%7Cfabio.bastianelli%40mail-bip.com%7C0bb5a42aac0c49cb876708d9f2c89510%7Cbb1a63ebeb09471aa00537b07792a5b5%7C0%7C0%7C637807765572787908%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&sdata=xWNOqSeS5JxDoBEWZ4aB63gLmsnTA8YGGfCOoLjo1eo%3D&reserved=0">https://smart-grid-use-cases.github.io/docs/usecases/platone/uc-it-2-congestion-management/</a>, <a href="https://platone-h2020.eu/data/deliverables/864300_M12_D1.1.pdf">https://platone-h2020.eu/data/deliverables/864300_M12_D1.1.pdf</a>)

H2020 Platone Italian Demonstrator Use Case 1-2 Market 2nd quarter 2022

H2020 Platone Italian Demonstrator Use Case 1-2 Market 1st quarter 2022

User Study Data for Paper "A case study in designing trustworthy interactions: implications for socially assistive robotics"

Experimental data collected for the user study described in Frontiers paper "A case study in designing trustworthy interactions: implications for socially assistive robotics" by Mengyu Zhong et al. Citation: Zhong, Mengyu, et al. "A case study in designing trustworthy interactions: implications for socially assistive robotics." Frontiers in Computer Science 5.1152532 (2023).

openCC (other)Jan 2023View details →

edi44/100

Weather data for the period 2009 to 2022 from the Open Field location at University Farms, Case Western Reserve University

Data from the Open Field weather station at University Farms of Case Western Reserve University include observations from 2009 to 2022. University Farms is located in Hunting Valley, Ohio. From 10/20/2009 to 10/30/2014, the weather station was located at N 41.496883, W 81.436117, when it was relocated to N 41.49759, W81.43738. Data include date/time (in 15-minute intervals), wind speed, wind gust speed, wind direction, air temperature, relative humidity, solar radiation, rainfall, soil moisture, soil temperatures at 0, 2, and 5 cm soil depth, and data logger battery charge.

edi44/100

Weather data for the period 2009 to 2022 from the North Woodlot location at University Farms, Case Western Reserve University

Data from the North Woodlot weather station at University Farms of Case Western Reserve University include observations from 2009 to 2022. University Farms is located in Hunting Valley, Ohio. The North Woodlot weather station is at N 41 29.969, W 81 25.234. Data include Date and time (in 15-minute intervals), wind speed, wind gust speed, air temperature, relative humidity, solar radiation, soil moisture, soil temperatures at 0, 2, and 5 cm soil depth, and data logger battery charge.

openCC (other)Jan 2023View details →

edi44/100

Termite casing data from the long-term Small Mammal Exclosure Study (SMES) at Jornada Basin LTER, 1995-2005

This data package contains termite activity data in plots with a range of herbivore exclusion treatments on Jornada Experimental Range (JER) and Chihuahuan Desert Rangeland Research Center (CDRRC) lands. Study sites were established in 1995; one in black grama grassland and the other in creosotebush shrubland to compare the impact of herbivores on ecosystem processes between these vegetation types. Parallel studies were established at the Sevilleta LTER site (New Mexico, USA) and Mapimi Biosphere Reserve (Durango, Mexico). Each study site is 1 km by 0.5 km in area. Four replicate experimental blocks were randomly located at each study site to measure vegetation responses using exclusion treatments including a) all mammalian herbivores, including cattle, lagomorphs, and rodents, b) lagomorphs and cattle only, c) cattle only, and d) control accessible to all herbivores. Thirty-six sampling points were positioned at 5.8-meter intervals on a systematically located 6 by 6 point grid within each plot. A permanent one-meter by one-meter vegetation measurement quadrat is located at each of the 36 points. Each spring and fall from 1995-2005, a tape measure was used to measure the length, diameter, and height in centimeters of each termite casing in these vegetation quadrats. This study is complete.

openCC (other)Aug 2019View details →

openother-openDec 2019View details →

Data and code for "Homeric -phi(n) is an oblique case marker"

This repository contains the data and R code for the paper "Homeric -phi(n) is an oblique case marker," Transactions of the Philological Society 118:343–375, 2020. doi: <a href="https://doi.org/10.1111/1467-968X.12192">10.1111/1467-968X.12192</a>.

opencc-by-4.0Jan 2020View details →

A set of six databases used in a study of the biogeography of Greater Caribbean reef fishes entitled: Comparing biodiversity databases: Greater Caribbean reef-fishes as a case study Iliana Chollett1, D. Ross Robertson2 1 Sea Cottage, Louisburgh, Co. Mayo, Ireland 2 Smithsonian Tropical Research Institute, Balboa, Panamá

A set of six databases used in a study of the biogeography of Greater Caribbean reef fishes entitled:   Comparing biodiversity databases: Greater Caribbean reef-fishes as a case study                                    Iliana Chollett, D. Ross Robertson     Database Authors: D Ross Robertson and Ernesto Peña, Smithsonian Tropical Research Institute, Panamá   This set of six databases contains georeferenced location records from six sources as described below.These six sources provided georeferenced records of occurrence of fishes found in the Greater Caribbean study area (6-330 N, 57-1000 W). Each occurrence record consists of a species name and associated latitude and longitude. Databases included in the comparisons made here are from five major online aggregators. Since their content overlaps to some extent, and OBIS, iDigBio and FishNet collaborate with GBIF, their data might be expected to produce similar biogeographic patterns. STRI includes a curated compendium of data from those five aggregators, enriched with data from many additional sources.   Only reef-associated fish species were included in the present analysis. These mostly represent demersal species known to occur on hard bottoms (coral, rock and oyster substrata), but also include species living on rubble, sand and vegetated bottoms within and around the immediate fringes of reefs, and pelagic species regularly found on reefs. All exotic and non-resident species and species other than reef-associated fishes were excluded from all databases prior to comparisons. Non-residents were defined as otherwise widespread species only rarely seen in the study area. Shore-fishes, including what are generally regarded as reef fishes, include those found in the waters of continental and insular shelves, i.e. between 0-200m. Reef-fish assemblages dominated by shallow-water taxa extend down to that depth in the study area (Baldwin et al. 2018). We used the shelf edge as a breakpoint and excluded records in areas deeper than 200m, identifying those areas using the General Bathymetric Chart of the Oceans (Kapoor, 1981; GEBCO Compilation Group, 2019).   Before the analyses, for all databases, duplicate records were deleted. Subsequently, records in the Pacific or on land were deleted. We used the Global Self-consistent, Hierarchical, High-resolution Geography Database (Wessel & Smith, 1996) to identify these areas. The spatial distribution of species-records in each database is shown in Figure 1 of the publication.   Global Biodiversity Information Facility (GBIF, https://www.gbif.org/): GBIF is an international network and research infrastructure aimed at providing open access to data about all types of life on earth. GBIF works through participant nodes using common standards and open-source tools that enable them to share information. Data from among the 49,000+ datasets hosted by GBIF that were used here range from those on museum specimens collected since the 18th century, to published scientific checklists, to curated  local checklists produced by trained science sources such as the Atlantic and Gulf Rapid Assessment Program (https://www.agrra.org/),to geotagged smartphone photos (that act as vouchers allowing verification) shared by amateur and scientific naturalists through iNaturalist (https://www.inaturalist.org/), to unvouchered, unverified and unverifiable observation records from untrained divers such as those contributing to DiveBoard (http://www.diveboard.com). GBIF data are standardized in Darwin Core format. GBIF data were obtained from a polygon of the region of study and subject to taxonomic review and selection after downloading. GBIF data were obtained from a polygon of the study area and subject to taxonomic review after downloading (accessed through the GBIF portal, https://www.gbif.org/, on or about 2019-05-19).   Ocean Biogeographic Information System (OBIS,  <a href="https://obis.org/">https://obis.org/</a>): OBIS is a global open-access data and information clearing-house on marine biodiversity (OBIS, 2019) that was adopted as a project of the Intergovernmental Oceanographic Data and Information Exchange of the Intergovernmental Commission of UNESCO . Its range of sources is similar to that of GBIF. OBIS hosts data from organizations or programs that join it as one of 13 “nodes”, and harvest the data from the IPT (Integrated Publishing Toolkit), where providers publish their data. The IPT is developed and maintained by the GBIF, and OBIS is a major contributor of marine data to GBIF. Data are standardized in Darwin Core format. OBIS data were obtained for the region of study by downloading data on each family, then retaining only data inside the study area, which were then subject to taxonomic review and selection (accessed through the OBIS portal, https://obis.org/, on or about 2019-05-19).   Integrated Digitized Biocollections (iDigBio, https://portal.idigbio.org/portal/search): iDigBio is sponsored by the a US National Science Foundation and run by the University of Florida that provides digital data from public, non-federal, US collections. Data are standardized in a Darwin Core format, and provided “as is”. IDigBio joined the GBIF network in 2017. IDigBio records were downloaded from a polygon of the region of study and subject to taxonomic review and selection (accessed through the iDigBio portal, https://portal.idigbio.org/portal/search, on or about 2019-05-19).   FishNet2 (http://www.fishnet2.net/): FishNet2 is a collaborative effort that aggregates data on fish collections around the world to share and distribute data on specimen holdings from ~75 museums, universities and other institutions. FishNet2 distributes data in Darwin Core, and data are provided “as is”. FishNet2 is part of the network VerNet, which has contributed to GBIF since 2013 and became part of IDigBio in 2016. While FishNet2 has made substantial efforts to georeference location-record data it hosts, many hosted records still lack georeferencing. FishNet2 data were obtained from a polygon of the study area and subject to taxonomic review after downloading (accessed through the Fishnet2 Portal, www.fishnet2.org, 2019-05-19).   FishBase (<a href="http://www.fishbase.org/">http://www.fishbase.org</a>): FishBase is a global biodiversity information system supervise by a consortium of nine non-USA international institutions, and hosts data on fin fishes and elasmobranchs  (Froese & Pauly, 2009). Information presented in FishBase is extracted from the scientific literature, reports and museum or aggregator (GBIF) databases, and standardized by a team of specialists. Data from Fishbase were downloaded for the following ecosystems: Caribbean Sea, Gulf of Mexico, Southeast U.S. Continental Shelf, Atlantic Ocean, Sargasso Sea and Bermuda, and subject to taxonomic review and selection after downloading (2019-05-19).   Smithsonian Tropical Research Institute (STRI; <a href="https://biogeodb.stri.si.edu/caribbean/en/pages">https://biogeodb.stri.si.edu/caribbean/en/pages</a>): The STRI database was compiled by DRR and Ernesto Peña at STRI’s Naos Marine Laboratory, and represents about 15 years accumulation of curated data (see below) from the following sources:  data downloaded at roughly two year intervals from the five aggregators; data from online databases of various museums that supply aggregators (data directly downloaded from a museum sometimes differs from that available in an aggregator from the same museum), including the Swedish Museum of Natural History, the American Museum of Natural History, the Natural History Museum of Denmark, the Gulf Coast Research Laboratory, the Colombian Museum of Natural Marine History, the United States National Museum, and the United States Geological Survey; data from national aggregators of Colombia (Sistema de Información Sobre Biodiversidad de Colombia (https://sibcolombia.net/), and  Sistema de Información Ambiental Marina de Colombia, https://siam.invemar.org.co/), Mexico (La Comisión Nacional para el Conocimiento y Uso de la Biodiversidad, CONABIO;   http://www.conabio.gob.mx/informacion/gis/), and Costa Rica (Museo de Zoologia de la Universidad de Costa Rica, http://museo.biologia.ucr.ac.cr/); verified (by DRR) underwater photographs of fishes taken at known locations; peer reviewed publications containing location information (species descriptions; taxonomic revisions of species, genera and families; regional and local checklists); fisheries reports; digital tagging data for species such as elasmobranchs; diving surveys and collections of local faunas by DRR (e.g. Robertson et al. 2019). In addition selected data from two sources that collect species lists at sites scattered throughout the Greater Caribbean are incorporated: from the Atlantic and Gulf Rapid Reef Assessment program (AGRRA, https://www.agrra.org/: Kramer & Lang, 2003) and from trained citizen scientists who contribute data on fishes to the Reef Environmental Education Foundation’s database (REEF: Pattengill-Semmens & Semmens, 2003). The bibliographic module (https://biogeodb.stri.si.edu/caribbean/en/library) of Robertson & VanTassel (2019) contains ~1700 publications linked to species names, among them the publications from which location data were extracted.   Data from the aggregators is presented “as is” and the aggregators themselves do not do data curation. Duplicates (and occasionally triplicates and quaduplicates) of the same museum record often are included from multiple sources (e.g. the original museum source, derivative checklists, an aggregator), sometimes with slightly different georeferenced coordinates. Data available in one year may subsequently disappear from an aggregator, and different data may be available for the same species under different names (e.g. the old and new names when a species is reassigned to another genus). Errors, sometimes large errors (Robertson, 2008), are common in aggregator data, from museums as well as other sources, and longstanding errors can seem to take on a perpetual existence. For example the damselfish Abudefduf saxatilis is a common and widespread inhabitant of tropical reefs on both sides of the Atlantic, but does not naturally occur outside that ocean. Despite the fact that its taxonomic status and range were resolved ~30 y ago (e.g. see Allen, 1991) museum data presented by the all five aggregators that contributed to the multi-source database used in this study currently (December 10, 2019) show large numbers of records of this species throughout the entire tropical Indo-Pacific, as well as across its native range in the Atlantic. Since many of the databases accumulating on aggregators are derivative (lists derived from records and from other derivative lists) it will become increasingly difficult to eliminate such errors as corrections to data in primary sources do not automatically propagate through the chain of usage by different databases. Due to increasing limitations on resources for taxonomic work, museums themselves have difficulty dealing with errors in specimen identity and location, and old specimens become unidentifiable, specimens never get returned when loaned out, or simply vanish, and entire collections can get destroyed by hurricanes or fires, or get dumped when museums close or experience a major change in mission. Georeferenced location data on fish distributions in the neotropics (and presumably most other areas) hosted by aggregators, particularly GBIF and OBIS, which take data from a broad range of source types, might best be described as messy, and the significant potential for errors in location records and an inability to verify records always needs to be taken into account when incorporating data from aggregators, primary museum sources, and analog sources.   Data considered for inclusion in the STRI database were screened as follows to exclude questionable records.  Data from two databases hosted by OBIS and GBIF were excluded entirely due to lack of reliability: BioGoMx (https://www.gulfbase.org/project/biodiversity-gulf-mexico-biogomx-database) and Diveboard (http://www.diveboard.com).  The only REEF data used were from “expert” REEF recorders on readily identifiable species that are unlikely to be confused with similar species (e.g. data for some genera of sparids, gerreids, labrisomids and gobies that include various sympatric species with very similar appearances, were not used).  After data from aggregators and museum sources were combined into a single database duplicate records were filtered out by rounding all records to three decimal places and eliminating duplicates, a process that inevitably deleted some valid records as well as duplicates. The sizes of the databases and abundance of such duplicates precluded individual manual exclusion. Finally, all location data for each species were revised by DRR by examining the distribution of its georeferenced coordinates overlayed on a digital map of the current known distribution range of that species (for such range information see Carpenter & De Angelis, 2002; Ebert et al., 2013; Last et al., 2016; Robertson & Van Tassell, 2019; IUCN Redlist species accounts for most species considered here: https://www.iucnredlist.org/search). Such revision took into account any recent modifications to taxonomy and distributions due to new data and new publications, or as a result of discussions between DRR and experts in the taxonomy of particular species or genera. Source information of many individual questionable records provided by aggregators with the hosted data was inspected to try and assess their validity. Records thought likely to be erroneous were deleted. Those included inexplicable records lacking adequate documentation located well outside the known distribution range, and records in unlikely habitats (e.g. on land for marine species; in deep water for shallow-water species). This revision process reduced the number of records by about 30%.   Data from the five individual aggregator databases that are used in the comparisons described here were all downloaded from their online portals during May, 2019. However, data from those five aggregators that were incorporated in the STRI database were downloaded in March 2017, with data from other sources described above added to the STRI database intermittently between then and May 2019, when the entire dataset was curated as described above. Hence the five individual aggregator databases analyzed in this study undoubtedly contain data not included in the version of the STRI database used in the present analyses.   Acknowledgements   Data acquisition and construction of the STRI database was supported by funds from STRI, the Smithsonian Marine Science Network, the Smithsonian Publications Fund, the Smithsonian’s Deep Reef Observation Project, the National Geographic Society, the IUCN Red List program, the Harte Research Institute, and CONABIO. We thank REEF and AGRRA for supplying species-location records, various people for taxonomic and location-record information used to construct that database (principal among them C Baldwin, S Brandl, K Conway B Frable, T Menut, T Munroe, R Robins, L Tornabene, J Van Tassell and B Victor), and hundreds of citizen-scientist submarine photographers whose images (see <a href="https://biogeodb.stri.si.edu/caribbean/en/contributors/citizen_scientists">https://biogeodb.stri.si.edu/caribbean/en/contributors/citizen_scientists</a>) acted as vouchers for location records.   References   Allen, G.R. (1991) Damselfishes of the World. Mergus, Melle, 271 p. Baldwin, C.C., Tornabene, L. & Robertson, D.R. (2018) Below the mesophotic. Scientific Reports, 8, 4920. Carpenter, K.E. (Ed) (2002) The living marine resources of the Western Central Atlantic. Vols 1-3, FAO, Rome, 2127 p. Ebert, D.A., Fowler, S., Compagno, L. (2013) Sharks of the World: a fully illustrated guide. Wild Nature Press, Plymouth. 528 p. GEBCO Compilation Group (2019) GEBCO 2019 Grid (doi:10.5285/836f016a-33be-6ddc-e053-6c86abc0788e). Kapoor, D.C. (1981) General bathymetric chart of the oceans (GEBCO). Marine Geodesy, 5, 73–80. Kramer, P.R. & Lang, J.C. (2003) Appendix one: The Atlantic and Gulf Rapid Reef Assessment (AGRRA) Protocols: Former Version 2. 2. Atoll Research Bulletin, 496, 611–624. Last, P. R., White, W.A., de Carvalho, M.R., Séret, B., Stehmann, F.W., & Naylor, J.P. (2016). Rays of the World. CSIRO, Clayton. 790 p. Pattengill-Semmens, C.V. & Semmens, B.X. (2003) Conservation and management applications of the reef volunteer fish monitoring program. Coastal Monitoring through Partnerships: Proceedings of the Fifth Symposium on the Environmental Monitoring and Assessment Program (EMAP) Pensacola Beach, FL, U.S.A., April 24–27, 2001 (ed. by B.D. Melzian), V. Engle), M. McAlister), S. Sandhu), and L.K. Eads), pp. 43–50. Springer Netherlands, Dordrecht. Robertson, D. R. (2008) Global biogeographic databases on marine fishes: caveat emptor. Diversity and Distributions, 14, 891-892 Robertson, D.R,, Dominguez-Dominguez, O., Lopez Arollo, Y.M., Moreno Mendoza. R., Simoes, N. (2019) Reef-associated fishes from the offshore reefs of western Campeche Bank, Mexico, with a discussion of mangroves and seagrass beds as nursery habitats. Zookeys 843: 71-115. <a href="https://doi.org/10.3897/zookeys.843.33873">https://doi.org/10.3897/zookeys.843.33873</a> Robertson, D.R & Van Tassell, J. (2019) Shorefishes of the Greater Caribbean: online information system. Version 2.0. Smithsonian Tropical Research Institute, Balboa, Panamá. <a href="https://biogeodb.stri.si.edu/caribbean/en/pages">https://biogeodb.stri.si.edu/caribbean/en/pages</a>. Wessel, P. & Smith, W.H.F. (1996) A global, self-consistent, hierarchical, high-resolution shoreline database. Journal of Geophysical Research: Solid Earth, 101, 8741–8743.

Fig. 29 in Systematics of the enigmatic South American Streblopus Van Lansberge, 1874 dung beetles and their transatlantic origin: a case study on the role of dispersal events in the biogeographical history of the Scarabaeinae (Coleoptera: Scarabaeidae)

Fig. 29. Current distribution of Streblopus van Lansberge, 1874 and the Old World groups with which it is believed to be more closely related plotted on an Upper Cretaceous palaeomap (~ 80 million years ago). Based particularly on the hypothesis in Tarasov & Génier (2015) that Streblopus is part of a clade otherwise composed uniquely of dung beetle lineages either exclusively distributed in Africa (Circellium, Chalconotus and Gyronotus) or with a distribution largely centred on that continent (Scarabaeini), and on the dating of the origin of the Scarabaeini as 71 million years ago (Gunter et al. 2016), we propose that the lineage that would eventually lead to Streblopus branched off from those groups in Africa some time between 95 and 71 million years ago, and that one of its descendent lineages (the only one living today) dispersed from its original continent to South America during the late Upper Cretaceous or the early Cenozoic. Since Africa and South America have not been connected by land since the Lower Cretaceous, the only way the ancestor of Streblopus could have reached South America was through transoceanic dispersal across the early South Atlantic. That dispersal probably happened by rafting on floating pieces of plants or other debris, as probably occurred with a large number of other organisms. Palaeomap modified from Scotese (2016); distribution area based on Balthasar (1963), Scholtz & Howden (1987), Davis et al. (2008) and our own results.

Fig. 14. Profemora. A‒B. Streblopus opatroides van Lansberge, 1874. A in Systematics of the enigmatic South American Streblopus Van Lansberge, 1874 dung beetles and their transatlantic origin: a case study on the role of dispersal events in the biogeographical history of the Scarabaeinae (Coleoptera: Scarabaeidae)

Fig. 14. Profemora. A‒B. Streblopus opatroides van Lansberge, 1874. A. ♂. B. ♀. C‒D. S. punctatus (Balthasar, 1938). C. ♂. D. ♀. Note the differences between the species and sexes in relation to the overall shape of the profemora and the presence of spurs on the anterior edge in males.

Fig. 12. Pronotum. A. Streblopus opatroides van Lansberge, 1874. B. S in Systematics of the enigmatic South American Streblopus Van Lansberge, 1874 dung beetles and their transatlantic origin: a case study on the role of dispersal events in the biogeographical history of the Scarabaeinae (Coleoptera: Scarabaeidae)

Fig. 12. Pronotum. A. Streblopus opatroides van Lansberge, 1874. B. S. punctatus (Balthasar, 1938). Note the differences in the umbilicate punctation and colour between the species.

Fig. 1. Streblopus opatroides van Lansberge, 1874. A‒B. Ordinary specimens, dorsal view. A in Systematics of the enigmatic South American Streblopus Van Lansberge, 1874 dung beetles and their transatlantic origin: a case study on the role of dispersal events in the biogeographical history of the Scarabaeinae (Coleoptera: Scarabaeidae)

Fig. 1. Streblopus opatroides van Lansberge, 1874. A‒B. Ordinary specimens, dorsal view. A. ♂. B. ♀. C‒D. Lectotype, ♂. C. Dorsal view. D. Attached labels.

Fig. 3 in Systematics of the enigmatic South American Streblopus Van Lansberge, 1874 dung beetles and their transatlantic origin: a case study on the role of dispersal events in the biogeographical history of the Scarabaeinae (Coleoptera: Scarabaeidae)

Fig. 3. Streblopus punctatus (Balthasar, 1938). A‒B. Holotype, ♀. A. Dorsal view. B. Attached labels. C‒D. Ordinary specimens, dorsal view. C. ♂. D. ♀.