Learning patterns of human behavior from sensor data is extremely important for high-level activity inference. We show how to extract and label a personās activities and signiļ¬...
Recommendations are crucial for the success of large websites. While there are many ways to determine recommendations, the relative quality of these recommenders depends on many f...
Parsing plays an important role in semantic role labeling (SRL) because most SRL systems infer semantic relations from 1-best parses. Therefore, parsing errors inevitably lead to ...
Traditional analysis methods for single-trial classiļ¬cation of electroencephalography (EEG) focus on two types of paradigms: phase locked methods, in which the amplitude of the ...
Christoforos Christoforou, Paul Sajda, Lucas C. Pa...
In information retrieval, the cluster hypothesis states: closely related documents tend to be relevant to the same request. We exploit this hypothesis directly by adjusting queryb...