Citation Prediction Using LDA Topic Models

This research investigates the use of deep learning models for predicting citation counts of academic research articles, leveraging Latent Dirichlet Allocation (LDA) scores as key features. By incorporating LDA-based topic modeling, the research explores the effectiveness of various deep learning models. The models are used to perform an analysis using the top 5 and top 10 LDA words. The results demonstrate that MLP and CNN outperform RNN models, with MLP consistently delivering the highest accuracy and lowest error rates.