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.
260
datasets available to search
ShareScore release 0.9.0
Dataset results
260 results for “Empirical Study”
Artifact for An Extensive Empirical Study of Nondeterministic Behavior in Static Analysis Tools
<p>This repository contains data for 'An Extensive Empirical Study of Nondeterministic Behavior in Static Analysis Tools' and the source code of the tool NDDetector that is used for performing the experiments in RQ2.</p><p>There are two directories, data and tool:</p><p><data> contains the data for the conclusion made in the two research questions, RQ1 and RQ2. (rq1 is Research Question 1s data)</p><p>In rq1/ there are:</p><p>final_results.csv - Contains 43 distinct results from 4 repositories (SOOT, WALA, FlowDroid, DroidSafe) that fix or report nondeterminism.</p><p>summary.pdf - Reports the number of nondeterminism results by tool repository at each stage of the qualitative study.</p><p>categorization.pdf - Reports the number of nondeterminism results by root cause categories at each component of analysis codebase in which the nondeterminism takes place</p><p>raw_data.zip - Contains the raw commits and issues extracted from 9 repositories (SOOT, DOOP, WALA, FlowDroid, DroidSafe, AmanDroid, TAJS, Code2Flow, PyCG)</p><p>key_words_results.zip - Contains the results extracted by each keyword (concurrency, concurrent, consistent, determinism, deterministic, different, flakiness, flaky, parallel, thread) from the raw data.</p><p>In rq2/ there are:</p><p>ICSE2024_AGGREGATE_DATA.csv - Contains the result distributions of each combination of target program, configuration hash, and tool aswell as the calculated consistency score.</p><p>analyze_results.py - Script that makes this data.</p><p>node_freqs - Contains the frequency of each node in the nondeterministic results we observed,it also keeps track of whether this particular node is a callee or caller or source/sink.</p><p>edge_dists - Contains the actual edge distributions of all of our results that behaved nondeterministically. it contains, for each result (edge/flow) across repetitions, which repetitions did or did not contain this edge/flow and which did. This means if you are interested in the actual differences across results generated by tool edge_dists/ is the place to look.</p><p>figure_8 - The raw data and occurences per node sheet for generating Figure_8.</p><p><tool> contains the framework and its source code that we used for conducting the experiments as well as the scripts that are used to post-process the detected nondeterminstic behavior and generate the summarized results in Section 4.</p>
The Classics Never Go Out of Style: An Empirical Study of Downgrades from the Bazel Build Technology
<p>This package contains the dataset used in our study: "The Classics Never Go Out of Style: An Empirical Study of Downgrades from the Bazel Build Technology."</p><p>For more information on understanding the folder structure, please read the <i>README</i>.md.</p>
An Empirical Comparative Study of Convolutional Neural Network and Support Vector Machine in Digital Signature for Digital Document Authentication
<p>Dataset dan figure of the research</p>
An Empirical Study of Next-Line Prediction in Build Systems Using CodeGen
<p>An Empirical Study of Next-Line Prediction in BuildSystems Using CodeGen.</p> <p>Build systems play a crucial role in software development and are responsible for compiling source code into executable programs. Despite their importance, build systems often receive limited attention because their impact is not directly visible to end users. This oversight can lead to inadequate maintenance, frequent build failures, and disruptions that require additional resources. Recognising and addressing the maintenance needs of build systems is essential to preventing costly disruptions and ensuring efficient software production.</p> <p>In this paper, we explore whether applying a Large Language Model (LLM) can reduce the burden of maintaining build systems. We aim to determine whether the prior content in build specifications provides sufficient context for an LLM to generate subsequent lines accurately. We conduct an empirical study on CodeGen, a state-of-the-art Large Language Model (LLM), using a dataset of 13,343 Maven build files. The dataset consists of the Expert dataset from the Apache Software Foundation (ASF) for fine-tuning (9,426 build files) and the Generalised dataset from GitHub for testing (3,917 build files). We observe that (i) fine-tuning on a small portion of data (i.e., 11\% of the fine-tuning dataset) provides the largest improvement in performance by 13.93\% (ii) When applied to the Generalised dataset, our fine-tuned model retains 83.86\% of its performance, indicating that it is not overfitted. Upon further investigation, we classify build code content into functional and metadata subgroups based on enclosing tags. Our fine-tuned model performs substantially better in suggesting functional than metadata build code. Our findings highlight the potential of leveraging LLMs like CodeGen to relieve the maintenance challenges associated with build systems, particularly in functional content. The study highlights the limitations of large language models in suggesting the metadata components of build code. Future research should focus on developing approaches to enhance the accuracy and effectiveness of metadata generation.</p> <p> </p> <p>Replication Package Structure:</p> <p>dataset.zip: Contains dataset for Expert and Generalised dataset.</p>
Understanding the Adoption of Modern JavaScript Features: An Empirical Study on Open-Source Systems
<p>This repository contains the data and analysis from an empirical study investigating the adoption trends of modern JavaScript features introduced with ECMAScript 6 (ES6) and beyond. By mining the source code history of 158 open-source JavaScript projects, the study identifies efforts to rejuvenate legacy code by replacing outdated constructs with modern ones. The findings highlight the extensive use of modern features, their widespread adoption within one to two years after ES6's release, and ongoing trends in the rejuvenation of JavaScript codebases.<br><br></p> <ul> <li> <p><strong>scripts.zip</strong>: Contains Python scripts used to analyze data and generate the graphs presented in the study's results.</p> </li> <li><strong>scripts-threats-analysis.zip</strong>: Contains the Python scripts used to analyze the projects without applying the study's filtering criteria and to generate the table presented in the Threats to Validity section.</li> <li> <p><strong>jsminer-tool.zip</strong>: Includes the tool developed to analyze GitHub repository history and collect metrics on the adoption of modern JavaScript features.</p> </li> <li> <p><strong>jsminer_database_backup.zip</strong>: Provides a PostgreSQL database dump containing all code review comments from the repositories analyzed in the study.</p> </li> </ul>
The Dataset of the PhD thesis titled "Operationalizing Values in Mobile Applications: A Mixed-Methods Empirical Study on Agriculture Apps for Bangladeshi Female Farmers"
<p>This package includes the survey (PVQ) questionnaire, demographic questions, focus groups questionnaire, interview questionnaire, and member checking summary used in this thesis.</p>
The dataset of the paper titled "Investigating End-Users' Values for Agriculture Mobile Applications Development: A Mixed-Methods Empirical Study on Bangladeshi Female Farmers"
<p>This package includes a survey questionnaire, demographic questions, focus groups questionnaire, interview questionnaire, and 10 main values with corresponding attributes and the referred names used in this paper submitted to IST.</p>
The dataset of the paper titled "Investigating End-Users' Values in Agriculture Mobile Applications Development: An Empirical Study on Bangladeshi Female Farmers"
<p>This package includes a survey questionnaire, demographic questions, and 10 main values with corresponding attributes and the referred names used in this paper submitted to JSS.</p>
Replication Package - How Do Requirements Evolve During Elicitation? An Empirical Study Combining Interviews and App Store Analysis
<p>This is the replication package for the paper titled "How Do Requirements Evolve During</p> <p>Elicitation? An Empirical Study Combining Interviews and App Store Analysis", by Alessio Ferrari, Paola Spoletini and Sourav Debnath.</p> <p> </p> <p>The package contains the following folders and files. </p> <p> </p> <p>**<strong>/Experiment Material</strong>**</p> <p>This folder contains the material used for the experiment, and provided to the participants.</p> <p>In particular, it includes the following files:</p> <p> </p> <p>- Happy CampingTM_briefdescription.pdf/docx: brief description of the product for which requirements need to be elicited</p> <p>- Hw_description.pdf/docx: desciption of the tasks to be performed by the participants</p> <p>- Modeling_Intro_Slides.pdf: introductory slides to modelling for requirements engineering</p> <p>- Self-assessment Questionnaire.pdf: first questionnaire to self-assess the mistakes, from the SaPeer method (https://doi.org/10.1007/s00766-020-00334-0) </p> <p>- Self-assessment Questionnaire (Second Interview).pdf: second questionnare to self-assess the mistakes, from the Sapeer method</p> <p> </p> <p>**<strong>/R-analysis</strong>**</p> <p> </p> <p>This is a folder containing all the R implementations of the the statistical tests included in the paper, together with the source .csv file used to produce the results. Each R file has the same title as the associated .csv file. The titles of the files reflect the RQs as they appear in the paper. The association between R files and Tables in the paper is as follows:</p> <p> </p> <p>- RQ1-1-analyse-story-rates.R: Tabe 1, user story rates </p> <p>- RQ1-1-analyse-role-rates.R: Table 1, role rates</p> <p>- RQ1-2-analyse-story-category-phase-1.R: Table 3, user story category rates in phase 1 compared to original rates</p> <p>- RQ1-2-analyse-role-category-phase-1.R: Table 5, role category rates in phase 1 compared to original rates</p> <p>- RQ2.1-analysis-app-store-rates-phase-2.R: Table 8, user story and role rates in phase 2</p> <p>- RQ2.2-analysis-percent-three-CAT-groups-ph1-ph2.R: Table 9, comparison of the categories of user stories in phase 1 and 2</p> <p>- RQ2.2-analysis-percent-two-CAT-roles-ph1-ph2.R: Table 10, comparison of the categories of roles in phase 1 and 2. </p> <p> </p> <p>The .csv files used for statistical tests are also used to produce boxplots. The association betwee boxplot figures and files is as follows. </p> <p> </p> <p>- RQ1-1-story-rates.csv: Figure 4 </p> <p>- RQ1-1-role-rates.csv: Figure 5</p> <p>- RQ1-2-categories-phase-1.csv: Figure 8</p> <p>- RQ1-2-role-category-phase-1.csv: Figure 9</p> <p>- RQ2-1-user-story-and-roles-phase-2.csv: Figure 13</p> <p>- RQ2.2-percent-three-CAT-groups-ph1-ph2.csv: Figure 14</p> <p>- RQ2.2-percent-two-CAT-roles-ph1-ph2.csv: Figure 17</p> <p>- IMG-only-RQ2.2-us-category-comparison-ph1-ph2.csv: Figure 15</p> <p>- IMG-only-RQ2.2-frequent-roles.csv: Figure 18</p> <p> </p> <p>NOTE: The last two .csv files do not have an associated statistical tests, but are used solely to produce boxplots.</p> <p> </p> <p>**<strong>/Data-Analysis</strong>**</p> <p> </p> <p>This folder contains all the data used to answer the research questions. </p> <p> </p> <p>**<strong>RQ1.xlsx</strong>**: includes all the data associated to RQ1 subquestions, two tabs for each subquestion (one for user stories and one for roles). The names of the tabs are self-explanatory of their content.</p> <p> </p> <p>**<strong>RQ2.1.xlsx</strong>**: includes all the data for the RQ1.1 subquestion. Specifically, it includes the following tabs:</p> <p> </p> <p>* Data Source-US-category: for each category of user story, and for each analyst, there are two lines. </p> <p>The first one reports the number of user stories in that category for phase 1, and the second one reports the</p> <p>number of user stories in that category for phase 2, considering the specific analyst. </p> <p> </p> <p>* Data Source-role: for each category of role, and for each analyst, there are two lines. </p> <p>The first one reports the number of user stories in that role for phase 1, and the second one reports the</p> <p>number of user stories in that role for phase 2, considering the specific analyst. </p> <p> </p> <p>* RQ2.1 rates: reports the final rates for RQ2.1. </p> <p>NOTE: The other tabs are used to support the computation of the final rates.</p> <p> </p> <p>**<strong>RQ2.2.xlsx</strong>**: includes all the data for the RQ2.2 subquestion. Specifically, it includes the following tabs:</p> <p> </p> <p>* Data Source-US-category: same as RQ2.1.xlsx</p> <p> </p> <p>* Data Source-role: same as RQ2.1.xlsx</p> <p> </p> <p>* RQ2.2-category-group: comparison between groups of categories in the different phases, used to produce Figure 14</p> <p> </p> <p>* RQ2.2-role-group: comparison between role groups in the different phases, used to produce Figure 17</p> <p> </p> <p>* RQ2.2-specific-roles-diff: difference between specific roles, used to produce Figure 18</p> <p> </p> <p>**<strong>NOTE:</strong>** the other tabs are used to support the computation of the values reported in the tabs above. </p> <p> </p> <p>**<strong>RQ2.2-single-US-category.xlsx</strong>**: includes the data for the RQ2.2 subquestion associated to single categories of user stories.</p> <p>A separate tab is used given the complexity of the computations. </p> <p> </p> <p>* Data Source-US-category: same as RQ2.1.xlsx</p> <p> </p> <p>* Totals: total number of user stories for each analyst in phase 1 and phase 2</p> <p> </p> <p>* Results-Rate-Comparison: difference between rates of user stories in phase 1 and phase 2, used to produce the file</p> <p>"img/IMG-only-RQ2.2-us-category-comparison-ph1-ph2.csv", which is in turn used to produce Figure 15</p> <p> </p> <p>* Results-Analysts: number of analysts using each novel category produced in phase 2, used to produce Figure 16.</p> <p>NOTE: the other tabs are used to support the computation of the values reported in the tabs above. </p> <p> </p> <p>**<strong>RQ2.3.xlsx</strong>**: includes the data for the RQ2.3 subquestion. Specifically, it includes the following tabs:</p> <p> </p> <p>* Data Source-US-category: same as RQ2.1.xlsx</p> <p> </p> <p>* Data Source-role: same as RQ2.1.xlsx</p> <p> </p> <p>* RQ2.3-categories: novel categories produced in phase 2, used to produce Figure 19</p> <p> </p> <p>* RQ2-3-most-frequent-categories: most frequent novel categories</p> <p> </p> <p>**<strong>/Raw-Data-Phase-I</strong>**</p> <p>The folder contains one Excel file for each analyst, s1.xlsx...s30.xlsx, plus the file of the original user stories with annotations (original-us.xlsx). Each file contains two tabs:</p> <p> </p> <p>- Evaluation: includes the annotation of the user stories as existing user story in the original categories (annotated with "E"), novel user story in a certain category (refinement, annotated with "N"), and novel user story in novel category (Name of the category in column "New Feature"). **<strong>NOTE 1:</strong>** It should be noticed that in the paper the case "refinement" is said to be annotated with "R" (instead of "N", as in the files) to make the paper clearer and easy to read. </p> <p> </p> <p>- Roles: roles used in the user stories, and count of the user stories belonging to a certain role.</p> <p> </p> <p>**<strong>/Raw-Data-Phaes-II</strong>**</p> <p>The folder contains one Excel file for each analyst, s1.xlsx...s30.xlsx. Each file contains two tabs:</p> <p> </p> <p>- Analysis: includes the annotation of the user stories as belonging to existing original </p> <p>category (X), or to categories introduced after interviews, or to categories introduced </p> <p>after app store inspired elicitation (name of category in "Cat. Created in PH1"), or to </p> <p>entirely novel categories (name of category in "New Category").</p> <p> </p> <p>- Roles: roles used in the user stories, and count of the user stories belonging to a certain role.</p> <p> </p> <p>**<strong>/Figures</strong>**</p> <p> </p> <p>This folder includes the figures reported in the paper. The boxplots are generated from the </p> <p>data using the tool http://shiny.chemgrid.org/boxplotr/. The histograms and other plots are </p> <p>produced with Excel, and are also reported in the excel files listed above. </p>
Not All Dependencies are Equal: An Empirical Study on Production Dependencies in NPM
<p>This dataset contains the data and scripts we used to conduct our study on production dependencies in npm. </p> <p>js_projects.zip contains 100 JavaScript projects mined from GitHub in November 2021. </p> <p>data_and_scripts.zip contains data on all installed dependencies from the 100 JavaScript projects and the scripts to process the data and obtain the results for our study. See the README file for more details. </p>
An Empirical Study of Package Management Issues via Stack Overflow
<p>The package manager (PM) is crucial to most technology stacks, acting as a broker to ensure that a verified dependency package is correctly installed, configured, or removed from an application. Diversity in technology stacks has led to dozens of PMs with various features. While our recent study indicates that package management features of PM are related to end-user experiences, it is unclear what those issues are and what information is required to resolve them. In this paper, we have investigated PM issues faced by end-users through an empirical study of content on Stack Overflow (SO). We carried out a qualitative analysis of 1,131 questions and their accepted answer posts for three popular PMs (i.e., Maven, npm, and NuGet ) to identify issue types, underlying causes, and their resolutions. Our results confirmed that end-users struggle with PM tool usage (approximately 64-72%). We observed that most issues are raised by end-users due to a lack of instructions and errors messages from PM tools. In terms of issue resolution, we observed that external link sharing is the most common practice to resolve PM issues. Additionally, we found that links pointing to useful resources (i.e., official documentation websites, tutorials, etc.) are the most frequently shared, indicating the potential for tool support and the ability to provide relevant information.</p>
Supplementary materials for "The values of public libraries: a systematic review of empirical studies of stakeholder perceptions"
<p>Supplementary materials for "The values of public libraries: a systematic review of empirical studies of stakeholder perceptions"</p> <p>https://www.emerald.com/insight/0022-0418.htm</p>
Generating Realistic Vulnerabilities via Neural Code Editing: An Empirical Study
<p>Using a commonly used synthetic dataset and one real-world dataset, we investigate the potential and gaps of three state-of-the-art neural code editors (Graph2Edit, Hoppity, SequenceR) for DL-based realistic vulnerability data generation, and two state-of-the-art vulnerability detectors (Devign, ReVeal) to evaluate the effectiveness of the generated realistic vulnerability data.</p> <p>Once the users have Docker installed download the Docker image "neural_editors_vulgen_docker.tar.xz".</p> <p>Then, check the README.md for detailed steps of reproducing the experiments.</p> <p>Besides, we also provide the simple package of the artifact "neural_editors_vulgen.zip". The raw data of our experiments is also provided in this simple package. However, using it to reproduce the experiments requires the users to set up the enviroments and dependencies for all the five tools, which is not recommanded.</p>
Dataset of the Paper "Architecture Decisions in AI-based Systems Development: An Empirical Study"
<p>This dataset collected from Stack Overflow (SO) and GitHub was used to conduct an empirical study on architecture decisions in AI-based systems development. We provide below a brief description of each file:</p> <p><strong>1. Dataset (SO).xlsx</strong></p> <p>contains the IDs and URLs of the labelled posts which are related to architecture decisions in AI-based systems development from SO, and the data extracted from these related SO posts.</p> <p><strong>2. Dataset (GitHub).xlsx</strong></p> <p>contains the project names, issue IDs, and issue URLs of AI-based projects selected from GitHub, and the data extracted from the relevant issues.</p> <p><strong>3. Extracted Data (SO+GitHub).xlsx</strong></p> <p>provides the final results of data extracted from SO posts and GitHub issues.</p>
On the Suitability of Hugging Face Hub for Empirical Studies
<p>This repository contains the data used in the paper title "On the Suitability of Hugging Face Hub for Empirical Studies". For RQ1 we share the survey responses and the interview transcription, while for RQ2 we share the link to the repository where the data is hosted.</p> <ul> <li>For RQ1, the survey responses are in an Excel file titled "Survey Responses Public.xlsx". The transcriptions of each interview are in a Word file titled "Transcription-intvw-slot-N.docx".</li> <li>For RQ2, we collected the data of the HFCommunity release of October 2023. It can be found in its <a href="https://som-research.github.io/HFCommunity/download.html">website</a>. We also share the <a href="../records/10020642">DOI of the dump</a>.</li> </ul>
Factors affecting onion production: an empirical study in the Raya Kobo District, Amhara Regional State, Ethiopia
<p>This is the excel data format covering farm household charactersitcs, plot characterstics and institutional characterstics and that support the findings of "<strong>Factors Affecting Onion Production: an Empirical Study in the Raya Kobo District, Amhara Regional State, Ethiopia".</strong></p> <p> </p>
Dataset related to: An Empirical Study on Low Code Programming using Traditional vs Large Language Model Support
<p>This repository contains data and prompts related to our research. The files included are:</p> <p><strong>prompt.py</strong>: This script contains the original prompts used in our study.</p> <p><strong>LLM_lowcode.mx20</strong>: This file includes posts and annotation data related to LLM-based low-code platforms.</p> <p><strong>Traditional_lowcode.mx20</strong>: This file includes posts and annotation data related to traditional low-code platforms.</p> <p>The .mx20 files can be opened using the MAXQDA software, which can be downloaded from the official website. MAXQDA offers a 14-day free trial.</p>
ATLAS+, Dataset for "An Empirical Study on Focal Methods in Deep-Learning-Based Approaches for Assertion Generation"
<p>Dataset for "An Empirical Study on Focal Methods in Deep-Learning-Based Approaches for Assertion Generation"</p>
Fairness in Diffusion Recommender Models: Insights from an Empirical Study of DiffRec
<p>Dataset for the experiments of the paper submission `Fairness in Diffusion Recommender Models: Insights from an Empirical Study of DiffRec`. The included datasets are Foursquare Tokyo (FKTY), MovieLens 1M with all attributes and users (ML1M).</p>
Scripts and data from the study 'Characterisation of magnetic atomic and molecular beamlines for the extraction of empirical scattering-matrices' (PCCP manuscript number CP-ART-04-2024-001785)
<p>This repository contains the data used for the article "Characterisation of magnetic atomic and molecular beamlines for the extraction of empirical scattering-matrices" and includes the code for generating the figures in the publication.</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.