In Studies in health technology and informatics ; h5-index 23.0
We present a work-in-progress software project which aims to assist cross-database medical research and knowledge acquisition from heterogeneous sources. Using a Natural Language Processing (NLP) model based on deep learning algorithms, topical similarities are detected, going beyond measures of connectivity via citation or database suggestion algorithms. A network is generated based on the NLP-similarities between them, and then presented within an explorable 3D environment. Our software will then generate a list of publications and datasets which pertain to a certain topic of interest, based on their level of similarity in terms of knowledge representation.
Schölly Reto, Yazijy Suhail, Kellmeyer Philipp
Data Mining (D057225), Information Services (D007255), Knowledge Bases (D051188)