A Novel Deep Neural Network-Based Approach to Measure Scholarly Research Dissemination Using Citations Network

We investigated the scientific research dissemination by analyzing publications and citation data, implying that not all citations are significantly important. Therefore, as alluded to existing state-of-the-art models employ feature-based techniques measure scholarly between multiple entities, our model implements convolutional neural network (CNN) with fastText-based pre-trained embedding vectors, utilizes only context its input distinguish important non-important citations. Moreover, we speculate using focal-loss class weight methods address inherited imbalance problems in classification datasets. Using a dataset of 10 K annotated contexts, achieved an accuracy 90.7% along 90.6% f1-score, case binary classification. Finally, present study comprehensiveness deployed on 3100 taken from ACL Anthology Reference Corpus. employed graph visualization open-source tool Gephi analyze various aspects graphs, for each respective behavior.