In this paper, we present a novel near-duplicate document detection method that can easily be tuned for a particular domain. Our method represents each document as a real-valued s...
Hannaneh Hajishirzi, Wen-tau Yih, Aleksander Kolcz
The traditional processing flow of segmentation followed by classification in computer vision assumes that the segmentation is able to successfully extract the object of interest....
We present several methods for mining knowledge from the query logs of the MSN search engine. Using the query logs, we build a time series for each query word or phrase (e.g., `Th...
Michail Vlachos, Christopher Meek, Zografoula Vage...
We present Hector, a software tool for combining different abstraction methods to extract sound models of heap-manipulating imperative programs with recursion. Extracted models ma...
Abstract. The assessment of Myocardial Blush Grade after primary angioplasty is a precious diagnostic tool to understand if the patient needs further medication or the use of speci...
Carlo Gatta, Juan Diego Gomez Valencia, Francesco ...