Biomedical images and captions are one of the major sources of information in online biomedical publications. They often contain the most important results to be reported, and pro...
Xin Chen, Caimei Lu, Yuan An, Palakorn Achananupar...
Hidden markov model (HMM) is frequently used for Pinyin-toChinese conversion. But it only captures the dependency with the preceding character. Higher order markov models can brin...
As a principled approach to capturing semantic relations of words in information retrieval, statistical translation models have been shown to outperform simple document language m...
Sentiment classification aims to automatically predict sentiment polarity (e.g., positive or negative) of users publishing sentiment data (e.g., reviews, blogs). Although traditio...
In this paper, we propose a new comprehensive methodology in order to evaluate the performance of noisy historical document recognition techniques. We aim to evaluate not only the...