Tag
This paper presents a benchmark for classifying research papers into good (highly cited) and non-good (retracted) using textual features from titles and abstracts, with models like SVM and neural networks achieving up to 91.12% accuracy.