Traditional classification methods assume that the training and the test data arise from the same underlying distribution. However, in several adversarial settings, the test set is...
This paper presents a multiagent architecture constructed for learning from the interaction between the atmosphere and the ocean. The ocean surface and the atmosphere exchange carb...
The scientific literature is a rich and challenging data source for research in systems biology, providing numerous interactions between biological entities. Text mining technique...
Assessment on collaborative student behavior is a longstanding issue in user modeling. Nowadays thanks to the proliferation of online learning and the vast amount of data on studen...
This paper introduces a new technique of document clustering based on frequent senses. The proposed system, GDClust (Graph-Based Document Clustering) works with frequent senses ra...