Network-based link prediction of scientific concepts -- a Science4Cast competition entry

TL;DR

This paper presents a network-based model for predicting scientific concept links, emphasizing the role of concept popularity and similarity, and incorporating temporal dynamics to enhance prediction accuracy in the Science4Cast competition.

Contribution

The study introduces a link prediction model that combines node degree, common neighbors, and temporal weighting, advancing methods for scientific knowledge network analysis.

Findings

High-degree nodes are more likely to form new links.

Similarity based on common neighbors improves prediction.

Temporal weighting of links enhances model performance.

Abstract

We report on a model built to predict links in a complex network of scientific concepts, in the context of the Science4Cast 2021 competition. We show that the network heavily favours linking nodes of high degree, indicating that new scientific connections are primarily made between popular concepts, which constitutes the main feature of our model. Besides this notion of popularity, we use a measure of similarity between nodes quantified by a normalized count of their common neighbours to improve the model. Finally, we show that the model can be further improved by considering a time-weighted adjacency matrix with both older and newer links having higher impact in the predictions, representing rooted concepts and state of the art research, respectively.