We present a new data set encoding localized semantics for 1014 images and a methodology for using this kind of data for recognition evaluation. This methodology establishes protoc...
Most traditional text clustering methods are based on "bag of words" (BOW) representation based on frequency statistics in a set of documents. BOW, however, ignores the ...
Jian Hu, Lujun Fang, Yang Cao, Hua-Jun Zeng, Hua L...
Thesaurus is a collection of words classified according to some relatedness measures among them. In this paper, we lay the theoretical foundations of thesaurus construction through...
With the increasing success and commercial integration of Volunteered Geographic Information (VGI), the focus shifts away from coverage to data quality and homogeneity. Within the ...
Extracting semantic relations among entities is an important first step in various tasks in Web mining and natural language processing such as information extraction, relation de...