"HistoryCR"

"We construct the dataset by targeting high-quality repositories that meet two rigorous criteria: a minimum of 150 stars and at least 3,000 Pull Requests (PRs). This selection process results in two distinct subsets:Java Subset: This subset encompasses 177,638 PRs, 832,592 code commits, and 18,392,959 code changes across 42 repositories, including 238,012 review comments. C++ Subset: Following the same rigorous selection process, this subset involves 233,350 PRs, 1,263,305 commits, and 18,736,357 code changes across 46 repositories, including 252,830 review comments, which provides a broader perspective for evaluating the cross-language scalability of HistoryCR. "