Who Benefits Most From Generative AI

Description

This dataset was developed to investigate workforce vulnerability, employee attrition risk, and the potential benefits of Generative Artificial Intelligence (GenAI) interventions in organizational settings. The study utilizes the IBM HR Analytics Employee Attrition dataset and extends it through machine learning, explainable artificial intelligence (XAI), and workforce segmentation techniques. The dataset contains 1,470 employee records and 35 original workforce-related variables, including demographic characteristics, job-related attributes, compensation factors, performance indicators, and employee satisfaction measures. To support the research objectives, several derived variables were created, including a Workforce Vulnerability Index, AI Benefit Score, AI Benefit Categories, Workforce Personas, and Priority AI Segments. The original data was taken from Pavansubhash. (2017). IBM HR Analytics Employee Attrition & Performance. https://www.kaggle.com/datasets/pavansubhasht/ibm-hr-analytics-attrition-dataset/data A Random Forest classifier was employed to model employee attrition risk, while SHAP (SHapley Additive exPlanations) analysis was used to identify the most influential factors contributing to workforce vulnerability. Based on these insights, a composite Vulnerability Index and AI Benefit Score were constructed to estimate which employee groups would benefit most from Generative AI-based workplace interventions. K-Means clustering was subsequently applied to identify distinct workforce personas representing different combinations of vulnerability, attrition risk, and AI augmentation potential. The final dataset includes original variables, engineered features, vulnerability measures, AI benefit indicators, and persona classifications. The repository also contains the complete Python code, analytical outputs, statistical validation results, and visualization files required to reproduce the findings reported in the associated research article. This dataset is intended for researchers, practitioners, and policymakers interested in workforce analytics, human augmentation, employee retention, explainable artificial intelligence, human resource management, and Generative AI adoption within organizations.

Steps to reproduce

Steps to Reproduce the Analysis Download all files contained in this repository and maintain the folder structure provided. Install Python 3.11 or later. Install the required Python libraries: pip install pandas numpy scikit-learn matplotlib seaborn scipy shap openpyxl Open the file "GenAI_Workforce_Analysis.py" in Jupyter Notebook, Jupyter Lab, VS Code, Spyder, or any Python IDE. Ensure that the dataset file "IBM_HR_Analytics_Employee_Attrition.csv" is located in the specified data directory and update the file path in the script if necessary. Execute the Python script sequentially. The script automatically performs: Data loading and preprocessing Feature engineering Employee attrition prediction using Random Forest SHAP explainability analysis Workforce Vulnerability Index creation AI Benefit Score development Workforce segmentation using K-Means clustering Workforce persona identification Statistical validation using Chi-Square and ANOVA tests Visualization generation Upon successful execution, the script creates an output directory named "GenAI_Workforce_Output". The following outputs are automatically generated: SHAP feature importance plot Workforce vulnerability by age group Workforce vulnerability by income group Job role vulnerability analysis Workforce segmentation matrix Persona comparison visualization AI benefit comparison visualization Excel workbook containing all analytical results Compare the generated outputs with the files provided in the Output folder to verify successful replication. Researchers may modify clustering parameters, machine learning algorithms, or scoring thresholds to extend the analysis and explore alternative workforce segmentation scenarios. Expected Results The analysis should reproduce four workforce personas: Stable Early-Career Workforce High-Risk AI Beneficiaries Moderate-Risk Growth Workforce Senior Low-Risk Professionals The analysis should further identify High-Risk AI Beneficiaries as the workforce segment with the highest vulnerability, highest attrition risk, and greatest potential benefit from Generative AI interventions. All figures, tables, and statistical results reported in the associated research article can be reproduced using the materials provided in this repository.

Institutions

• Aditya University

  Andhra Pradesh, Sūrampālem

Categories

Licence