In this paper, we address the question of what kind of knowledge is generally transferable from unlabeled text. We suggest and analyze the semantic correlation of words as a gener...
The classical probabilistic models attempt to capture the Ad hoc information retrieval problem within a rigorous probabilistic framework. It has long been recognized that the prim...
- We address the issues of improving the feature generation methods for the value-function approximation and the state space approximation. We focus the improvement of feature gene...
Abstract. Geographic Information Retrieval (GIR) is a specialized Information Retrieval (IR) branch that deals with information related to geographical locations. Traditional IR en...
Categorizing web-based videos is an important yet challenging task. The difficulties arise from large data diversity within a category, lack of labeled data, and degradation of vi...