An analysis of intelligent methods and algorithms for unlabeled data processing

Intelligent algorithms and method is well-suited to many problems in data processing, where unlabeled data may be abundant. We survey previously used selection strategies for intelligent model, and propose two novel algorithms to address their shortcomings, focus on Active Learning (AL). While has already been shown to markedly reduce the annotation efforts for many sequence labeling tasks compared to random selection, AL remains unconcerned about the internal structure of the selected sequences (typically, sentences). We propose a semi-supervised AL approach for sequence labeling.