Discovering Research Hypotheses in Social Science Using Knowledge Graph Embeddings

In an era of ever-increasing scientific publications available, scientists struggle to keep pace with the literature, interpret research results and identify new hypotheses falsify. This is particularly in fields such as social sciences, where automated support for discovery still widely unavailable unimplemented. this work, we introduce system that supports identifying hypotheses. With idea knowledge graphs help modeling domain-specific information, machine learning can be used most relevant facts therein, frame problem hypothesis a link prediction task, ComplEx model predict relationships between entities graph representing papers their experimental details. The final output consists fully formulated including newly discovered triples (hypothesis statement), along supporting statements from evidence history). A quantitative qualitative evaluation carried using experts field. Encouraging show simple combination methods serve basis discovery.