Anticipating the availability of large questionanswer datasets, we propose a principled, datadriven Instance-Based approach to Question Answering. Most question answering systems ...
This paper introduced the four tracks that WIM-Lab Fudan University had taken part in at TREC 2007. For spam track, a multi-centre model was proposed considering the characteristi...
Jun Xu, Jing Yao, Jiaqian Zheng, Qi Sun, Junyu Niu
The performance of parallel query processing in a cluster of index servers is crucial for modern web search systems. In such a scenario, the response time basically depends on the...
Claudine Santos Badue, Ricardo A. Baeza-Yates, Ber...
Large-scale cluster-based Internet services often host partitioned datasets to provide incremental scalability. The aggregation of results produced from multiple partitions is a f...
Abstract. As grid sizes increase, the need for self-organization and dynamic reconfigurations is becoming more and more important, and therefore the convergence of grid computing ...