Find research datasets worth reusing

Stalization of the Furuta pendulum: Linear, nonlinear and AI-based controllers (perturbed case)

Stabilization control of the Furuta pendulum in Matlab/Simulink Simscape environmet in perturbed case. Linear controllers <ul> <li>LQR</li> <li>PID</li> </ul> Nonlinear controllers <ul> <li>Feedback Linearization</li> <li>SMC</li> </ul> AI-based controllers <ul> <li>Feedback Linearization with adaptive nerual networks</li> <li>Reiforcement Learning</li> <li>Feedback Linearization with Reinforcement Learning compensation</li> </ul>

Survey on attitudes and knowledge about AI in health professionals and students (Spanish Version)

The Gebauer and Eckert1 survey on attitudes and knowledge about artificial intelligence was used, a translation into Spanish was made and adaptation of the survey to the Cuban context was made, and then it was validated by experts.   1Gebauer S, Eckert C. Survey of US physicians’ attitudes and knowledge of AI. BMJ Evid Based Med [Internet]. 2024 Feb 14 [cited 2024 May 7]; Available from: <a href="https://doi.org/10.1136/bmjebm-2023-112726">https://doi.org/10.1136/bmjebm-2023-112726</a>

Dataset S4-6: Leveraging co-evolutionary insights and AI-based structural modeling to unravel receptor-peptide ligand-binding mechanisms

<h3>Significance statement:</h3> This study presents proof-of-concept for a rapid and inexpensive alternative to classical structure-based approaches for resolving ligand-receptor binding mechanisms. It relies on a multilayered bioinformatic approach that leverages genomic data across diverse species in combination with AI-based structural modeling to identify true ligand and receptor homologues, and subsequently predict their binding mechanisms. In silico findings were validated by multiple experimental approaches, which investigated the effect of amino acid changes in the proposed binding pockets on ligand-binding, complex formation with a co-receptor essential for downstream signaling, and activation of downstream signaling. Our analysis combining evolutionary insights, in silico modeling and functional validation provides a framework for structure-function analysis of other peptide-receptor pairs, which could be easily implemented by most laboratories. <h3>Zip file contains:</h3> Dataset S4: Plasmid maps of constructs used in this study. Dataset S5: AFM and AF3 predicted structures (.pdb) and AFM confidence metrics (.pae) Dataset S6: Unedited files (.tiff) of co-IP and western blotting.

opencc-by-4.0May 2024View details →

Diversity and Inclusion (D&I)- Related AI Incidents Repository

This is a repository of Diversity and Inclusion (D&I) related AI incidents. This repository is proposed in our recently submitted paper titled "AI for All: Identifying AI incidents Related to Diversity and Inclusion".

dryad32/100

Generative AI enhances individual creativity but reduces the collective diversity of novel content

Creativity is core to being human. Generative AI—made readily available by powerful large language models (LLMs)—holds promise for humans to be more creative by offering new ideas, or less creative by anchoring on generative AI ideas. We study the causal impact of generative AI ideas on the production of short stories in an online experiment where some writers obtained story ideas from an LLM. We find that access to generative AI ideas causes stories to be evaluated as more creative, better written, and more enjoyable, especially among less creative writers. However, generative AI-enabled stories are more similar to each other than stories by humans alone. These results point to an increase in individual creativity at the risk of losing collective novelty. This dynamic resembles a social dilemma: with generative AI, writers are individually better off, but collectively a narrower scope of novel content is produced. Our results have implications for researchers, policy-makers, and practitioners interested in bolstering creativity.

opencc-zeroJun 2024View details →

AI-BASED INSPECTION OF RECYCLED CARBON FIBRE FABRIC CLEAN SAMPLES

This dataset consists of different non defective samples of carbon fibers. The defective samples and their classes for these carbon fiber types are give <a href="../records/11203952">here</a>.  cleandataset_samples.pkl - It is a pickle file. This contains the dataset for clean samples. Since these are non defective samples there are no class labels. This dataset consists of approximately 10 samples per carbon fiber matrix type.  load_cleandata.py - This is the sample code to load the data. Explanation to the shapes and the structure of the data are present here. The dataset presented has received funding from the European Union’s Horizon Europe research and innovation program under grant agreement No 101057394. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union. Neither the European Union nor the granting authority can be held responsible for them.

2024 ACNS AI Tutorial Unsupervised Learning Dataset

Data needed for a tutorial on unsupervised learning for the 2024 ACNS conference. Associated github repository is: <a title="ACNS AI Tutorial Repo" href="https://github.com/CAMM-UTK/acns-AI-tutorial.git">https://github.com/CAMM-UTK/acns-AI-tutorial</a>

IoT Sensor based Sports Activity Monitoring using Skill Inheritance and Optimization based on AI

Open the record for dataset details and reuse information.

Validation and Test Datasets for "High-resolution AI image dataset for diagnosing oral submucous fibrosis and squamous cell carcinoma"

This deposition contains the validation and test dataset for our study "High-resolution AI image dataset for diagnosing oral submucous fibrosis and squamous cell carcinoma". The training dataset for this study can be found at the following DOI: [10.5281/zenodo.12636426].

AI and Political Marketing: How Indonesian React to The Presidential Candidate AI-Generated Advertisement

The integration of Artificial Intelligence (AI) into political campaigns, particularly through social media, has revolutionized voter behavior influencing mechanisms. This research explores the impact of AI-generated content on public decision-making in the context of the 2024 Indonesian presidential elections, focusing on the Prabowo-Gibran pair. Drawing upon selective exposure theory, the study investigates how perceived source slant, perceived source bias, and perceived effects of opponents' news diet influence political participation likelihood. Through a quantitative survey of 150 Indonesian citizens active on social media, the study reveals significant positive correlations between perceived source slant, bias, and political participation likelihood. The findings underscore the pivotal role of AI-generated content in shaping political engagement and suggest implications for enhancing voter mobilization strategies in the digital age.

opencc-by-4.0May 2024View details →

Evaluation of literature reviews & primary sources thereof regarding the continuous development of AI

Replication Package for the Paper The Past, Present, and Future of Research on the Continuous Development of AI. The Replication package includes: <ul> <li>evaluation_literatureStudy.html/ipynb: <ul> <li>Analysis to answer RQ1: Analysis regarding the literature reviews and their demographic information and empirical standards</li> <li>Analysis to answer RQ2: main topics covered by systematic literature reviews and calculation of trends over time</li> <li>Analysis to answer RQ3: BerTopic Analysis for clustering primary sources</li> </ul> </li> <li>input.zip: Input Datasets for analysis</li> <li>generatedOutput: Output generated by analysis in evaluation_literatureStudy.ipynb - for further information of each csv or pdf file please refer to the analysis itself</li> <li>methodology.zp:  <ul> <li>SLRFORSLR-data-includedSources - Excluded Sources.csv: Papers which were excluded because they did not fulfill inclusion/exclusion criteria</li> <li>SLRFORSLR-data-includedSources - Search terms.csv: database specific search terms and amount of results from each search term</li> <li>ThreatsToValidity-DeviationOfPrimarySources.pdf: As discussed in the threats to validity the identified literature studies where the primary sources differ to the stated amount by the literature reviews' authors.</li> <li>BerTopic_Models.zip: trained BerTopic models used to cluster primary sources</li> </ul> </li> </ul>

The Future of Journalism: AI Acceptance and It's Implication in Newsroom

The integration of Artificial Intelligence (AI) in journalism marks a significant evolution in news reporting and the media landscape, while at the same time introducing changes at least on three levels: the role of journalists in text production, their replacement in executing some activities, and interaction with the audience. This study proposes a conceptual model based on the Technology Acceptance Model (TAM) and the Unified Theory of Acceptance and Use of Technology (UTAUT) to understand the acceptance of AI in Newsrooms in Indonesia. From the measurement results, a total of 100 respondents were obtained for the study. The acceptance was examined using linear regression analysis, with the TAM Model and UTAUT Model serving as the theoretical framework and the following independent variables: Perceived Ease of Use (PE) and Usefulness (UT) and Effort Expectancy (SAP). From the findings, it's clear that various factors positively influence the behavior intention (BI) in adopting AI within the newsroom context. To effectively adopt AI, selecting technologies that prioritize user benefits and performance enhancement is crucial. Decision-makers recognize AI's potential to significantly aid journalistic work in the future, indicating a growing acceptance and integration of AI technologies in newsroom operations.

opencc-by-sa-4.0Jul 2024View details →

Exploration of humor in generative AI

Twelve conversations with different open-source LLM-based chatbots (Cohere4AI, Gemma, Llama, Mistral) exploring humor understanding and production with zero-shot and n-shot strategies. Conversations were conducted via HuggingFace chat in Spring 2024, exported as HTML pages and bundled in a zip file.

Supplementary materials for an AI-based method for estiamting the potential runout distance of post-seismic debris flows

We included several figures and tables in this file to support the study on runout distance estimation using a AI-based method, mainly focusing on the precipitation downscaling, calibration, assessment, and rainfall threshold calculation. Additionally, the prediction results of all the debris flow catchments were included in a table when intraday rainfall ranges from 40 to 100 mm.

Dataset of the Paper "Exploring the Problems, their Causes and Solutions of AI Pair Programming: A Study on GitHub and Stack Overflow"

The dataset collected from GitHub Discussions, GitHub Issues, and Stack Overflow is used to conduct an empirical study on the problems, causes, and solutions of using GitHub Copilot in practice. A brief description of each document in the dataset is provided below: 1. Dataset(GitHub_Discussions).xlsx contains the Discussion IDs and URLs in the Copilot category of GitHub Discussions, and the data extracted from the related discussions along with analysis results. 2. Dataset(GitHub_Issues).xlsx contains the Issue IDs and URLs of the labelled issues which are related to Copilot from GitHub Issues, and the data extracted from the related issues along with analysis results. 3. Dataset(SO_Posts).xlsx contains the SO Post IDs and URLs of the labelled posts which are related to Copilot from Stack Overflow, and the data extracted from the related posts along with analysis results. 4. Extracted_Data.xlsx contains the final results of the data extracted from GitHub Discussions, GitHub Issues, and SO posts. 5. pilot labelling folder contains three .xlsx files (i.e., Pilot_Labelling(GitHub_Discussions).xlsx, Pilot_Labelling(GitHub_Issues), and Pilot_Labelling(SO)), with each file corresponding to one of the three data sources and containing the pilot data labelling results with the Cohen's kappa value.

2000-2002 Dataset [1/7] for the models trained and tested in the paper 'Can AI be enabled to dynamical downscaling? Training a Latent Diffusion Model to mimic km-scale COSMO-CLM downscaling of ERA5 over Italy'

This repository contains part 1/7 of the full dataset used for the models of the <a href="https://arxiv.org/abs/2406.13627">preprint</a> "Can AI be enabled to dynamical downscaling? Training a Latent Diffusion Model to mimic km-scale COSMO-CLM downscaling of ERA5 over Italy".  This dataset comprises 3 years of normalized hourly data for both low-resolution predictors [16 km] and high-resolution target variables [2km] (2mT and 10-m U and V), from 2000-2002. Low-resolution data are preprocessed ERA5 data while high-resolution data are preprocessed VHR-REA CMCC data. Details on the performed preprocessing are available in the paper. This part 1/7 of the dataset also includes files related to metadata, static data, normalization, and plotting. To use the data, clone the corresponding <a href="https://github.com/DSIP-FBK/DiffScaler">repository</a>, unzip this zip file in the data folder, and download from Zenodo the other parts of the dataset listed in the related works.

2009-2011 Dataset [4/7] for the models trained and tested in the paper 'Can AI be enabled to dynamical downscaling? Training a Latent Diffusion Model to mimic km-scale COSMO-CLM downscaling of ERA5 over Italy'

This repository contains part 4/7 of the full dataset used for the models of the <a href="https://arxiv.org/abs/2406.13627">preprint</a> "Can AI be enabled to dynamical downscaling? Training a Latent Diffusion Model to mimic km-scale COSMO-CLM downscaling of ERA5 over Italy".  This dataset comprises 3 years of normalized hourly data for both low-resolution predictors [16 km] and high-resolution target variables [2km] (2mT and 10-m U and V), from 2009-2011. Low-resolution data are preprocessed ERA5 data while high-resolution data are preprocessed VHR-REA CMCC data. Details on the performed preprocessing are available in the paper. To use the data, clone the corresponding <a href="https://github.com/DSIP-FBK/DiffScaler">repository</a>, unzip this zip file in the data folder, and download from Zenodo the other parts of the dataset listed in the related works.

2012-2014 Dataset [5/7] for the models trained and tested in the paper 'Can AI be enabled to dynamical downscaling? Training a Latent Diffusion Model to mimic km-scale COSMO-CLM downscaling of ERA5 over Italy'

This repository contains part 5/7 of the full dataset used for the models of the <a href="https://arxiv.org/abs/2406.13627">preprint</a> "Can AI be enabled to dynamical downscaling? Training a Latent Diffusion Model to mimic km-scale COSMO-CLM downscaling of ERA5 over Italy".  This dataset comprises 3 years of normalized hourly data for both low-resolution predictors [16 km] and high-resolution target variables [2km] (2mT and 10-m U and V), from 2012-2014. Low-resolution data are preprocessed ERA5 data while high-resolution data are preprocessed VHR-REA CMCC data. Details on the performed preprocessing are available in the paper. To use the data, clone the corresponding <a href="https://github.com/DSIP-FBK/DiffScaler">repository</a>, unzip this zip file in the data folder, and download from Zenodo the other parts of the dataset listed in the related works.

2015-2017 Dataset [6/7] for the models trained and tested in the paper 'Can AI be enabled to dynamical downscaling? Training a Latent Diffusion Model to mimic km-scale COSMO-CLM downscaling of ERA5 over Italy'

This repository contains part 6/7 of the full dataset used for the models of the <a href="https://arxiv.org/abs/2406.13627">preprint</a> "Can AI be enabled to dynamical downscaling? Training a Latent Diffusion Model to mimic km-scale COSMO-CLM downscaling of ERA5 over Italy".  This dataset comprises 3 years of normalized hourly data for both low-resolution predictors [16 km] and high-resolution target variables [2km] (2mT and 10-m U and V), from 2015-2017. Low-resolution data are preprocessed ERA5 data while high-resolution data are preprocessed VHR-REA CMCC data. Details on the performed preprocessing are available in the paper. To use the data, clone the corresponding <a href="https://github.com/DSIP-FBK/DiffScaler">repository</a>, unzip this zip file in the data folder, and download from Zenodo the other parts of the dataset listed in the related works.

2006-2008 Dataset [3/7] for the models trained and tested in the paper 'Can AI be enabled to dynamical downscaling? Training a Latent Diffusion Model to mimic km-scale COSMO-CLM downscaling of ERA5 over Italy'

This repository contains part 3/7 of the full dataset used for the models of the <a href="https://arxiv.org/abs/2406.13627">preprint</a> "Can AI be enabled to dynamical downscaling? Training a Latent Diffusion Model to mimic km-scale COSMO-CLM downscaling of ERA5 over Italy".  This dataset comprises 3 years of normalized hourly data for both low-resolution predictors [16 km] and high-resolution target variables [2km] (2mT and 10-m U and V), from 2006-2008. Low-resolution data are preprocessed ERA5 data while high-resolution data are preprocessed VHR-REA CMCC data. Details on the performed preprocessing are available in the paper. To use the data, clone the corresponding <a href="https://github.com/DSIP-FBK/DiffScaler">repository</a>, unzip this zip file in the data folder, and download from Zenodo the other parts of the dataset listed in the related works.