Hangul

Abstract: A reduced keyboard disambiguating system for the Korean language using word-level disambiguation to resolve ambiguities in keystrokes. A plurality of letters are assigned to each of a plurality of data keys, so that keystrokes on these keys are ambiguous. A user may enter a keystroke sequence wherein each keystroke corresponds to the entry of one letter of a word. Because individual keystrokes are ambiguous, the keystroke sequence can potentially match more than one word with the same number of letters. The keystroke sequence is processed by matching the input keystroke sequence to corresponding stored words or other interpretations. The 14 consonant jamos of the Korean alphabet are assigned to the data keys in their standard alphabetical sequence. The 10 vowel jamos are likewise assigned to the same set of data keys in their standard alphabetical sequence, such that a majority of the data keys are assigned both one or more consonant jamos and one or more vowel jamos. The result is a reduced keyboard that is easily understood and quickly learned by native speakers of Korean, and that is efficient for purposes of disambiguating textual interpretations of input sequences of ambiguous keystrokes.

Abstract: This is the first volume of its kind to treat specifically the critical role of language in Korean culture and society Early on, contributors examine the invention and use of the Korean alphabet, South Korea's "standard language" vs North Korea's "cultured language," and Korean in contact with Chinese and Japanese Several topics representative of Korean socio-cultural vocabulary are discussed, followed by a consideration of Korean honorifics and other related issues Two chapters on Korean media follow Finally, contributors look at salient features of the language, narrative structure, and dialectal variation All chapters are accompanied by a set of student questions and a useful bibliography A beginning level of proficiency in Korean is sufficient to digest the Korean examples with facility, making this volume accessible to a wide range of students

Abstract: In two separate studies, 100 South Korean kindergartners and 100 second graders were administered tests of speed of processing and phonological-processing skills as well as a Korean Hangul reading test. Speed of processing tasks were significantly correlated with most of the reading-related tasks but not with Hangul reading itself. Across studies, only tests of syllable and phoneme awareness (either onset or both onset and coda), as well as naming speed, were significantly associated with Korean word-reading. In both studies, in regression equations including all measured skills and age, only syllable and phoneme awareness uniquely predicted Hangul word recognition. Results underscore the uniqueness of the Korean Hangul orthography, which requires children to be sensitive to both syllable- and phoneme-level linguistic and orthographic units.

Abstract: In spite of the advances in recognition technology, handwritten Hangul recognition (HHR) remains largely unsolved due to the presence of many confusing characters and excessive cursiveness in Hangul handwritings. Even the best existing recognizers do not lead to satisfactory performance for practical applications and have much lower performance than those developed for Chinese or alphanumeric characters. To improve the performance of HHR, here we developed a new type of recognizers based on deep neural networks (DNNs). DNN has recently shown excellent performance in many pattern recognition and machine learning problems, but have not been attempted for HHR. We built our Hangul recognizers based on deep convolutional neural networks and proposed several novel techniques to improve the performance and training speed of the networks. We systematically evaluated the performance of our recognizers on two public Hangul image databases, SERI95a and PE92. Using our framework, we achieved a recognition rate of 95.96 % on SERI95a and 92.92 % on PE92. Compared with the previous best records of 93.71 % on SERI95a and 87.70 % on PE92, our results yielded improvements of 2.25 and 5.22 %, respectively. These improvements lead to error reduction rates of 35.71 % on SERI95a and 42.44 % on PE92, relative to the previous lowest error rates. Such improvement fills a significant portion of the large gap between practical requirement and the actual performance of Hangul recognizers.