In this paper we will briefly describe the approaches taken by the Berkeley Cheshire Group for the GikiCLEF task of the QA track. Because the task was intended to model some aspec...
Previously we proposed a scheme to generate fuzzy rule-based multiclassification systems by means of bagging, mutual information-based feature selection, and a multicriteria geneti...
Krzysztof Trawinski, Arnaud Quirin, Oscar Cord&oac...
We present a framework for tracking rigid objects based on an adaptive Bayesian recognition technique that incorporates dependencies between object features. At each frame we fin...
Our experiments in the Robust track this year focused on predicting query difficulty and using this prediction for improving information retrieval. We developed two prediction alg...
Elad Yom-Tov, Shai Fine, David Carmel, Adam Darlow...
Bursty features in text streams are very useful in many text mining applications. Most existing studies detect bursty features based purely on term frequency changes without takin...
Wayne Xin Zhao, Jing Jiang, Jing He, Dongdong Shan...