• Journal of KIISE
• /
• v.41 no.9
• /
• pp.642-651
• /
• 2014

This paper proposes three probabilistic models for syllable-based Korean morphological analysis, and presents the performance of proposed probabilistic models. Probabilities for the models are acquired from POS-tagged corpus. The result of 10-fold cross-validation experiments shows that 98.3% answer inclusion rate is achieved when trained with Sejong POS-tagged corpus of 10 million eojeols. In our models, POS tags are assigned to each syllable before spelling recovery and morpheme generation, which enables more efficient morphological analysis than the previous probabilistic models where spelling recovery is performed at the first stage. This efficiency gains the speed-up of morphological analysis. Experiments show that morphological analysis is performed at the rate of 147K eojeols per second, which is almost 174 times faster than the previous probabilistic models for Korean morphology.

• Phonetics and Speech Sciences
• /
• v.3 no.3
• /
• pp.69-75
• /
• 2011

This paper investigates phrase positional effects on the timing of F0 (pitch) peaks in Tokyo Japanese disyllabic words with varying accent type (HL or LH) and phrase position (final or non final). The F0 peak timing was normalized by the total word duration ('normalized H timing'). The normalized H timing was significantly affected by accent type and phrase position. The H timing was later in the LH accent type than in the HL accent type, and in non final positions than in final positions. In addition, to examine the validity of the quantitative results, different models of phrase position effects were compared by measuring H timing in two approaches: normalization versus relative distance measures. For the normalization measures, the H timing was measured as the time of the F0 peak divided by the total word duration or by the duration of the tone bearing syllable. For the relative distance measures, the H timing was measured as the distance in milliseconds from the end of the word or from the end of the associated syllable. The best model was the normalization by the total word duration, rather than by the duration of the tone bearing syllable. This means that phrase positional effects on the timing of F0 peaks in Japanese disyllabic words are best modeled in terms of proportion of the total word duration.

• MALSORI
• /
• no.21_24
• /
• pp.90-110
• /
• 1992

The purpose of this thesis is to investigate the intonation features of the last two syllables of rhythmic units, with the exception of the sentence final unit, in the Seoul dialect of Korean. The Model 5500 Sona-graph was used to measure the pitch and duration of the target syllables. There are two classes of materials. One class was designed to determine the intonation of rhythmic units in a natural situation and the other to investigate the intonation of rhythmic units in an artificial situation, in which speakers were asked to read the materials pausing only at the marked boundaries, with a view to identifying the intonation of Seoul dialect more clearly. The findings of this investigation are as follows: (1) Korean averages an 11% rising intonation between the two syllables at the end of a rhythmic unit. (2) The rising rate between the final two syllables' pitch values at the subject rhythmic unit is generally higher than those at other units in a sentence and it seems to be meaningful syntactically. (3) Before a boundary the rhythmic units undergo 'pre-lowering', in which the pitch gradually lowers from the first syllable to the penultimate. (4) Every syllable in each rhythmic unit tends to lengthen when speakers read the materials with a pause between units and the tendency is most salient at the final syllable before a boundary.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.33B no.8
• /
• pp.105-113
• /
• 1996

In this paper, a new korean digita speech recognition technique was proposed using muktolayer perceptron (MLP). In spite of its weakness in dynamic signal recognition, MLP was adapted for this model, cecause korean syllable could give static features. It is so simle in its structure and fast in its computing that MLP was used to the suggested system. MLP's input vectors was transformed using karhunen-loeve transformation (KLT), which compress signal successfully without losin gits separateness, but its physical properties is changed. Because the suggested technique could extract static features while it is not affected from the changes of syllable lengths, it is effectively useful for korean numeric recognition system. Without decreasing classification rates, we can save the time and memory size for computation using KLT. The proposed feature extraction technique extracts same size of features form the tow same parts, front and end of a syllable. This technique makes frames, where features are extracted, using unique size of windows. It could be applied for continuous speech recognition that was not easy for the normal neural network recognition system.

• The Journal of the Acoustical Society of Korea
• /
• v.21 no.3
• /
• pp.274-284
• /
• 2002

Word segmentation errors occurring in text preprocessing often insert incorrect words into recognition vocabulary and cause poor language models for Korean large vocabulary continuous speech recognition. We propose an automatic word segmentation algorithm using Markov chains and syllable-based n-gram language models in order to correct word segmentation error in teat corpora. We assume that a sentence is generated from a Markov chain. Spaces and non-space characters are generated on self-transitions and other transitions of the Markov chain, respectively Then word segmentation of the sentence is obtained by finding the maximum likelihood path using syllable n-gram scores. In experimental results, the algorithm showed 91.58% word accuracy and 96.69% syllable accuracy for word segmentation of 254 sentence newspaper columns without any spaces. The algorithm improved the word accuracy from 91.00% to 96.27% for word segmentation correction at line breaks and yielded the decomposition accuracy of 96.22% for compound-noun decomposition.

• Journal of KIISE:Software and Applications
• /
• v.37 no.8
• /
• pp.653-658
• /
• 2010

In this paper, we propose a word spacing system which can be performed with just a small memory. We focus on significant memory reduction while maintaining the performance of the system as much as the latest studies. Our proposed method is based on the theory of Hidden Markov Model. We use only probability information not adding any rule information. Two types of features are employed: 1) the first features are the spacing patterns dependent on each individual syllable and 2) the second features are the values of transition probability between the two syllable-patterns. In our experiment using only the first type of features, we achieved a high accuracy of more than 91% while reducing the memory by 53% compared with other systems developed for mobile application. When we used both types of features, we achieved an outstanding accuracy of more than 94% while reducing the memory by 76% compared with other system which employs bigram syllables as its features.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.17 no.4
• /
• pp.199-208
• /
• 2022

Korean is an agglutinative language, and one or more morphemes are combined to form a single word. Part-of-speech tagging method separates each morpheme from a word and attaches a part-of-speech tag. In this study, we propose a new Korean part-of-speech tagging method based on the Head-Tail tokenization technique that divides a word into a lexical morpheme part and a grammatical morpheme part without decomposing compound words. In this method, the Head-Tail is divided by the syllable boundary without restoring irregular deformation or abbreviated syllables. Korean part-of-speech tagger was implemented using the Head-Tail tokenization and deep learning technique. In order to solve the problem that a large number of complex tags are generated due to the segmented tags and the tagging accuracy is low, we reduced the number of tags to a complex tag composed of large classification tags, and as a result, we improved the tagging accuracy. The performance of the Head-Tail part-of-speech tagger was experimented by using BERT, syllable bigram, and subword bigram embedding, and both syllable bigram and subword bigram embedding showed improvement in performance compared to general BERT. Part-of-speech tagging was performed by integrating the Head-Tail tokenization model and the simplified part-of-speech tagging model, achieving 98.99% word unit accuracy and 99.08% token unit accuracy. As a result of the experiment, it was found that the performance of part-of-speech tagging improved when the maximum token length was limited to twice the number of words.

• Journal of Korea Society of Industrial Information Systems
• /
• v.29 no.3
• /
• pp.13-22
• /
• 2024

Currently, with the proliferation of digital devices, the significance of handwritten texts in daily lives is gradually diminishing. As the use of keyboards and touch screens increase, a decline in Korean handwriting quality is being observed across a broad spectrum of Korean documents, from young students to adults. However, Korean handwriting still remains necessary for many documentations, as it retains individual unique features while ensuring readability. To this end, this paper aims to implement an application designed to improve and correct the quality of handwritten Korean script The implemented application utilizes the CRAFT (Character-Region Awareness For Text Detection) model for handwriting area detection and employs the VGG-Feature-Extraction as a deep learning model for learning features of the handwritten script. Simultaneously, the application presents the user's handwritten Korean script's reliability on a syllable-by-syllable basis as a recognition rate and also suggests the most similar fonts among candidate fonts. Furthermore, through various experiments, it can be confirmed that the proposed application provides an excellent recognition rate comparable to conventional commercial character recognition OCR systems.

• The Transactions of the Korea Information Processing Society
• /
• v.13 no.10
• /
• pp.553-558
• /
• 2024

Morphological analysis involves segmenting morphemes, the smallest units of meaning or grammatical function in a language, and assigning part-of-speech tags to each morpheme. It plays a critical role in various natural language processing tasks, such as named entity recognition and dependency parsing. Much of modern natural language processing relies on deep learning-based language models, and Korean morphological analysis can be broadly categorized into sequence-to-sequence methods and sequential labeling methods. This study proposes a morphological analysis approach using the transformer encoder for sequential labeling to perform syllable-level part-of-speech tagging, followed by morpheme restoration and tagging through a pre-analyzed dictionary. Additionally, the CBOW method was used to extract syllable-level embeddings in lower dimensions, designing a lightweight morphological analyzer model with reduced parameters. The proposed model achieves fast inference speed and low parameter usage, making it efficient for use in resource-constrained environments.

• Proceedings of the Korean Society for Language and Information Conference
• /
• 2002.02a
• /
• pp.259-270
• /
• 2002

I examine various controversial aspects of Chinese prosody-tone structure, syllable structure, stress, and intonation-and stress the need to view all of these as interacting systems, aspects of a hierarchical prosodic structure. 1 examine various proposals at these various levels of the hierarchy and suggest which are most appropriate. Specifically, 1 suggest the adoption of Bao's version of syllable and tone, and Chen's account of stress. As for intonation, it is still not possible to make any definitive claims regarding an optimal model, but I examine work done by Kratochvil, Shih, and Carding et al, and suggest promising directions for future work.