Early Prediction of Scientific Impact Based on Multi-Bibliographic Features and Convolutional Neural Network

Abstract

The increasingly available large-scale bibliographic data that generate a heterogeneous network provide opportunities to detect, track, and predict the evolution of science. Recently, many efforts have been devoted to quantifying the impact of scientific papers within different citation time windows. However, the complex patterns of the citation network make it difficult to predict future citations on the basis of a short time window. Accordingly, we present a data-centric methodology to predict long-term scientific impact by combining numerous bibliographic features and convolutional neural network. More specifically, we first expand the input features from the annual citation records to the features of the whole heterogeneous bibliographic information network that completely represents the topology structure of academic activities. Then, a convolutional neural network model is designed to capture the complex nonlinear relationships between the early network features and the final cumulative citation count. Finally, we conduct an experiment on papers of Markov Chain from 1980 to 1985. The result shows that the prediction performance can be improved by 5% to baseline models under the same problem definition and with the same dataset. Meanwhile, the long-term scientific impacts are strongly correlated with its recognition by authoritative authors or venues in the early stage. INDEX TERMSScientific impact, heterogeneous bibliographic information network, convolutional neural network, citation count.

Deep Learning Techniques for Community Detection in Social Networks

“…(15) Set the weight of node v i in the next iteration plus the value of temp. (16) end for (17) end for (18) end while (19) Return the ID of the node with a highest weight.…”

A Survey on the Recent Advances of Deep Community Detection

“…Generally, the CNN and its variants have been widely adopted in network embedding [52][53][54][55]. For example, Xu et al [52] re-formatted the complex network topology adjacent matrix into an image and designed a CNN model to extract and classify relevant features.…”

Utilizing Citation Network Structure to Predict Citation Counts: A Deep Learning Approach

“…[14] extracted six article features, two journal features, eight reference features, nine author features and five early citation features, and then fed them into a multi-layer BP neural network to predict the five-year citations of the paper; [24] used Recurrent Neural Networks(RNNs) to take the counts of citation of the paper each year as input, and the output is the number of citation counts of the paper in the next K years; [25] developed the GRU-CPM model based on the GRU network (a specific form of Recurrent Neural Networks), which predicts the amount of citation counts by extracting the text features and author features of the paper, and obtained higher prediction accuracy; [43] proposed a complex deep learning model to predict citation counts by combining the peer-reviewed text of the paper with other features. [44] used the features of heterogeneous bibliographic networks and convolutional neural networks (CNN) to predict the number of citations of each paper in ten years, and improved the prediction accuracy by 5% compared to the baseline models.…”