Predicting research trends with semantic and neural networks with an application in quantum physics

Code for the paper: https://www.pnas.org/content/early/2020/01/13/1914370116/tab-article-info / https://arxiv.org/abs/1906.06843

A followup competition in the domain of Machine Learning and Aritificial Intelligence: Science4Cast

by Mario Krenn ([email protected]) and Anton Zeilinger ([email protected])

We show an implementation in MATLAB and in python for creating a semantic network from data, and making predictions into the future.

CreateAndAnalyseSEMNET

Contains a simple example, how to start from APS and arXiv files to create a Semantic Networks, calculate network theoretic properties, and make predictions

PersonalizedSuggestions

Shows how to make personalized surprising predictions from personal papers, SemNet and the neural network predictions

FullSemNet

Full concept list with 6372 entries, and the Semantic Network which is 6372x6372 entries with time information about concept emergence and connection emergence