http://dl.acm.org/citation.cfm?id=2701119 http://dl.acm.org/citation.cfm?id=2701119 Fri, 19 Dec 2014 00:00:00 GMT 2701119 http://dl.acm.org/citation.cfm?id=2629545
This article presents an approach to nonnative pronunciation variants modeling and prediction. The pronunciation variants prediction method was developed by generalized transformation-based error-driven learning (GTBL). The modified goodness of pronunciation (GOP) score was applied to effective mispronunciation detection using logistic regression machine learning under the pronunciation variants prediction. English-read speech data uttered by Korean-speaking learners of English were collected, then pronunciation variation knowledge was extracted from the differences between the canonical phonemes and the actual phonemes of the speech data. With this knowledge, an error-driven learning approach was designed that automatically learns phoneme variation rules from phoneme-level transcriptions. The learned rules generate an extended recognition network to detect mispronunciations.]]> Fri, 19 Dec 2014 00:00:00 GMT 2629545 http://dl.acm.org/citation.cfm?id=2637478
In statistical machine translation, the standard methods such as MERT tune a single weight with regard to a given development data. However, these methods suffer from two problems due to the diversity and uneven distribution of source sentences. First, their performance is highly dependent on the choice of a development set, which may lead to an unstable performance for testing. Second, the sentence level translation quality is not assured since tuning is performed on the document level rather than on sentence level. In contrast with the standard global training in which a single weight is learned, we propose novel local training methods to address these two problems.]]> Fri, 19 Dec 2014 00:00:00 GMT 2637478 http://dl.acm.org/citation.cfm?id=2629575
Ancient language manuscripts constitute a key part of the cultural heritage of mankind. As one of the most important languages, Chinese historical calligraphy work has contributed to not only the Chinese cultural heritage but also the world civilization at large, especially for Asia. To support deeper and more convenient appreciation of Chinese calligraphy works, based on our previous work on the probabilistic retrieval of historical Chinese calligraphic character manuscripts repositories, we propose a system framework of the multi-feature-based Chinese calligraphic character images probabilistic retrieval in the mobile cloud network environment, which is called the DPRC. To ensure retrieval efficiency, we further propose four enabling techniques: (1) DRL-based probability propagation, (2) optimal data placement scheme, (3) adaptive data robust transmission algorithm, and (4) index support filtering scheme.]]> Fri, 19 Dec 2014 00:00:00 GMT 2629575