In this paper, we propose a framework of hypotheses generation from scientific literature. The framework mainly consists of three modules: ontology-based domain knowledge representation; represent the existing research themes with association rule; and inference generation. The hypotheses generation process is accompanied by interactive selection of researchers. This provides an environment for researchers to assist hypotheses generation by discovering inferences from current research themes in scientific literature. Based on their research interest and background knowledge, researchers can evaluate the reasonability and utility of the inferences, and select some of them as their future research directions.
展开▼