MOOC Renewal Dataset and ILP Optimization Code

This dataset accompanies the research article “Optimizing MOOC Update Planning: A Multi-Objective ILP Approach Based on the Next Release Problem.” It contains all input data, scripts, and solver outputs used to reproduce the Integer Linear Programming (ILP) experiments described in the study. The dataset integrates learner survey data, candidate update parameters, Python optimization scripts, and solver-generated outputs for different effort budgets and weighting schemes. All materials are anonymized and released under a CC BY 4.0 license to support transparency and reproducibility. Contents Candidate_Updates_Table.csvThe normalized dataset of 26 MOOC update candidates extracted from learner survey responses (n = 437) for the Introduction to Cyber Security course on SWAYAM. Columns: Update, Learner Value (0–1), Effort Units (1–3), Category, Dependencies Categories: Content, Assessment, Engagement, Support MOOC_ILP_Normalized_v2.zipContains the primary Colab-ready ILP implementation (ILP_MOOC_Optimization.ipynb) and intermediate processed data. Implements multi-objective ILP with normalized learner value, effort, and coverage-based diversity Benchmarks against a greedy heuristic Generates summary tables and figures for analysis MOOC_ILP_outputs.zipIncludes all solver outputs and visualizations generated by the experiments: summary_metrics.csv – aggregated normalized performance scores selected_updates.csv – selected updates per budget and scheme fig_budget_vs_Z1.png and fig_Z1_vs_Dcov_balanced.png – graphical summaries code.docxA readable documentation version of the Colab code with inline comments describing preprocessing, ILP formulation, and export steps for replication and review. Technical Summary Software: Python 3.10, PuLP 3.3.0, Pandas 2.2.2, Matplotlib 3.9.0 Solver: CBC (default open-source linear solver) Optimization Objectives: Maximize learner value (normalized) Minimize total effort (normalized inverse) Maximize diversity coverage across update categories Budgets tested: 8, 10, 12, 14, and 16 effort units.Weighting schemes: Value-focused (0.7/0.2/0.1), Balanced (0.5/0.25/0.25), and Diversity-focused (0.4/0.2/0.4).