• Journal of the Korean Society for Library and Information Science
• /
• v.39 no.1
• /
• pp.45-58
• /
• 2005

This Paper is a comparative study of feature selection methods for Korean web documents clustering. First, we focused on how the term feature and the co-link of web documents affect clustering performance. We clustered web documents by native term feature, co-link and both, and compared the output results with the originally allocated category. And we selected term features for each category using $X^2$, Information Gain (IG), and Mutual Information (MI) from training documents, and applied these features to other experimental documents. In addition we suggested a new method named Max Feature Selection, which selects terms that have the maximum count for a category in each experimental document, and applied $X^2$ (or MI or IG) values to each term instead of term frequency of documents, and clustered them. In the results, $X^2$ shows a better performance than IG or MI, but the difference appears to be slight. But when we applied the Max Feature Selection Method, the clustering Performance improved notably. Max Feature Selection is a simple but effective means of feature space reduction and shows powerful performance for Korean web document clustering.

• Phonetics and Speech Sciences
• /
• v.12 no.1
• /
• pp.33-42
• /
• 2020

This study identified the substitution of /u/ for /o/ (e.g., pyəllo [pyəllu]) in Korean based on the speech corpus as a function of grammar items. Korean /o/ and /u/ share the vowel feature [+rounded], but are distinguished in terms of tongue height. However, researchers have reported that the merger of Korean /o/ and /u/ is in progress, making them indistinguishable. Thus, in this study, the frequency of the phonetic manifestation /u/ of the underlying form of /o/ for each grammar item was calculated in The Korean Corpus of Spontaneous Speech (Seoul Corpus 2015) which is a large corpus from a total of 40 speakers from Seoul or Gyeonggi-do. It was then confirmed that linking endings, particles, and adverbs ending with /o/ in the word-final position were substituted for /u/ approximately 50% of the stimuli, whereas, in nominal items, they were replaced at a frequency of less than 5%. The high rates of substitution were the special particle "-do[du]" (59.6%) and the linking ending "-go[gu]" (43.5%) among high-frequency items. Observing Korean pronunciation in real life provides deep insight into its theoretical implications in terms of speech recognition.

• The Journal of Korean Medical History
• /
• v.28 no.2
• /
• pp.37-45
• /
• 2015

Objectives : A quantitative analysis on the vocabulary in the English translation version of Donguibogam. Methods : This study quantitatively analyzed the English-translated texts of Donguibogam with the Corpus-based analysis, and compared the quantitative results analyzing the texts of original Donguibogam. Results : As the results from conducting the corpus analysis on the English-translation version of Donguibogam, it was found that the number of total words (Token) was about 1,207,376, and the all types of used words were about 20.495 and the TTR (Type/Token Rate) was 1.69. The accumulation rate reaching to the high-ranking 1000 words was 83.54%, and the accumulation rate reaching to the high-ranking 2000 words was 90.82%. As the words having the high-ranking frequency, the function words like 'the, and of, is' mainly appeared, and for the content words, the words like 'randix, qi, rhizoma and water' were appeared in multi frequencies. As the results from comparing them with the corpus analysis results of original version of Donguibogam, it was found that the TTR was higher in the English translation version than that of original version. The compositions of function words and contents words having high-ranking frequencies were similar between the English translation version and the original version of Donguibogam. The both versions were also similar in that their statements in the parts of 'Remedies' and 'Acupuncture' showed higher composition rate of contents words than the rate of function words. Conclusions : The vocabulary in the English translation version of Donguibogam showed that this book was a book keeping the complete form of sentence and an Korean medical book at the same time. Meanwhile, the English translation version of Donguibogam had some problems like the unification of vocabulary due to several translators, and the incomplete delivery of word's meanings from the Chinese character-culture area to the English-culture area, and these problems are considered as the matters to be considered in a work translating Korean old medical books in English.

• Cross-Cultural Studies
• /
• v.25
• /
• pp.627-648
• /
• 2011

The study analyzed a diachronic distribution, social meanings and social evaluations of ROOJIN. 'Headline Database of Newspaper Articles' and 'Balanced Corpus of Contemporary Written Japanese' by KOKKEN were used as research data. There were 305 newspaper articles (About 0.2%) which contained the word ROOJIN at 'Headline Database of Newspaper Articles'. The number of newspaper articles related to ROOJIN started to increase in a rapid rate in 1972 and 1973. They were also increased in 1976, from 1981 to 1987, 1992 and 1993. The reasons of increasing of newspaper articles related to ROOJIN on those 4 periods of time could be summarized as follows. Firstly, there was a increasement of ROOJIN who are lonely, are not able to move about freely or live alone. Secondly, the understanding of a symptom of aging called BOKE was necessary. Thirdly, there were negative evaluations in a society towards ROOJIN. There were 453 cases which contained the word ROOJIN at 'Balanced Corpus of Contemporary Written Japanese' on the data since 2000. The most frequently used words were ones that are related to senior care facilities. There were 109 cases (24%) which contain those words. '~SISETSU', '~SENTA-', '~HO-MU' were presented as words related to senior care facilities. Among them, 78 cases contained the word '~HO-MU' which was similar to a home with family members. The second most frequently used words were ones that are related to 'welfare for the aged' and they are led by 'medical care for the aged'. They occupied about 8%. Institutionalization of medical care for the aged, medical expenses, nursing were presented as words related to 'medical care for the aged'. Words that were related to 'welfare for the aged' led by 'senior care facilities' and 'medical care for the aged' occupied about 32% of research data. As mentioned above, problems of the aged in Modern Japan such as negative evaluations in a society towards ROOJIN, ROOJIN who are lonely, are not able to move about freely or live alone, BOKE could be identified by analyzing the data. Also, The frequent usage of words such as 'Home for the aged', 'medical care for the aged' and 'nursing' could be identified. The outcome of analysis suggested that a family traditionally had a function of solving problems of the aged but that function was reduced in modern Japan. It also suggested that there was a tendency to outsource problems of the aged as much as possible.

• Phonetics and Speech Sciences
• /
• v.11 no.2
• /
• pp.33-44
• /
• 2019

This paper dealt with how realization of pauses in utterance is affected by speech style, gender, and generation. For this purpose, we analyzed the frequency and duration of pauses. Pauses were categorized into four types: pause with breath, pause with no breath, utterance medial pause, and utterance final pause. Forty-eight subjects living in Seoul were chosen from the Korean Standard Speech Database. All subjects engaged in reading and spontaneous speech, through which we could also compare the realization between the two speech styles. The results showed that utterance final pauses had longer durations than utterance medial pauses. It means that utterance final pause has a function that signals the end of an utterance to the audience. For difference between tasks, spontaneous speech had longer and more frequent pauses because of cognitive reasons. With regard to gender variables, women produced shorter and less frequent pauses. For male speakers, the duration of pauses with breath was significantly longer. Finally, for generation variable, older speakers produced more frequent pauses. In addition, the results showed several interaction effects. Male speakers produced longer pauses, but this gender effect was more prominent at the utterance final position.

• KIPS Transactions on Software and Data Engineering
• /
• v.8 no.3
• /
• pp.115-122
• /
• 2019

For Korean phoneme recognition, Hidden Markov-Gaussian Mixture model(HMM-GMM) or hybrid models which combine artificial neural network with HMM have been mainly used. However, current approach has limitations in that such models require force-aligned corpus training data that is manually annotated by experts. Recently, researchers used neural network based phoneme recognition model which combines recurrent neural network(RNN)-based structure with connectionist temporal classification(CTC) algorithm to overcome the problem of obtaining manually annotated training data. Yet, in terms of implementation, these RNN-based models have another difficulty in that the amount of data gets larger as the structure gets more sophisticated. This problem of large data size is particularly problematic in the Korean language, which lacks refined corpora. In this study, we introduce CTC algorithm that does not require force-alignment to create a Korean phoneme recognition model. Specifically, the phoneme recognition model is based on convolutional neural network(CNN) which requires relatively small amount of data and can be trained faster when compared to RNN based models. We present the results from two different experiments and a resulting best performing phoneme recognition model which distinguishes 49 Korean phonemes. The best performing phoneme recognition model combines CNN with 3hop Bidirectional LSTM with the final Phoneme Error Rate(PER) at 3.26. The PER is a considerable improvement compared to existing Korean phoneme recognition models that report PER ranging from 10 to 12.

• The Korean Journal of Archival Studies
• /
• no.70
• /
• pp.85-117
• /
• 2021

The purpose of this study is to investigate and analyze the current status of unit tasks, unit task operation, and record management problems used by local governments, and to present improvement measures using text-based big data technology based on the implications derived from the process. Local governments are in a serious state of record management operation due to errors in preservation period due to misclassification of unit tasks, inability to identify types of overcommon and institutional affairs, errors in unit tasks, errors in name, referenceable standards, and tools. However, the number of unit tasks is about 720,000, which cannot be effectively controlled due to excessive quantities, and thus strict and controllable tools and standards are needed. In order to solve these problems, this study developed a system that applies text-based analysis tools such as corpus and tokenization technology during big data analysis, and applied them to the names and construction terms constituting the record management standard. These unit task operation support tools are expected to contribute significantly to record management tasks as they can support standard operability such as uniform preservation period, identification of delegated office records, control of duplicate and similar unit task creation, and common tasks. Therefore, if the big data analysis methodology can be linked to BRM and RMS in the future, it is expected that the quality of the record management standard work will increase.

• Journal of the Korean Society for information Management
• /
• v.38 no.2
• /
• pp.153-172
• /
• 2021

This study is to create a deep learning based classification model to examine the characteristics of panic disorder and to classify the panic disorder tendency literature by the panic disorder corpus constructed for the present study. For this purpose, 5,884 documents of the panic disorder corpus collected from social media were directly annotated based on the mental disease diagnosis manual and were classified into panic disorder-prone and non-panic-disorder documents. Then, TF-IDF scores were calculated and word co-occurrence analysis was performed to analyze the lexical characteristics of the corpus. In addition, the co-occurrence between the symptom frequency measurement and the annotated symptom was calculated to analyze the characteristics of panic disorder symptoms and the relationship between symptoms. We also conducted the performance evaluation for a deep learning based classification model. Three pre-trained models, BERT multi-lingual, KoBERT, and KcBERT, were adopted for classification model, and KcBERT showed the best performance among them. This study demonstrated that it can help early diagnosis and treatment of people suffering from related symptoms by examining the characteristics of panic disorder and expand the field of mental illness research to social media.

• The Journal of the Acoustical Society of Korea
• /
• v.41 no.3
• /
• pp.359-366
• /
• 2022

In this paper, we present the Korean menu-ordering Sentence Text-to-Speech (TTS) system using conformer-based FastSpeech2. Conformer is the convolution-augmented transformer, which was originally proposed in Speech Recognition. Combining two different structures, the Conformer extracts better local and global features. It comprises two half Feed Forward module at the front and the end, sandwiching the Multi-Head Self-Attention module and Convolution module. We introduce the Conformer in Korean TTS, as we know it works well in Korean Speech Recognition. For comparison between transformer-based TTS model and Conformer-based one, we train FastSpeech2 and Conformer-based FastSpeech2. We collected a phoneme-balanced data set and used this for training our models. This corpus comprises not only general conversation, but also menu-ordering conversation consisting mainly of loanwords. This data set is the solution to the current Korean TTS model's degradation in loanwords. As a result of generating a synthesized sound using ParallelWave Gan, the Conformer-based FastSpeech2 achieved superior performance of MOS 4.04. We confirm that the model performance improved when the same structure was changed from transformer to Conformer in the Korean TTS.

• Journal of Life Science
• /
• v.33 no.8
• /
• pp.663-679
• /
• 2023

This review analyzes research trends related to new drug development using artificial intelligence from 2010 to 2022. This analysis organized the abstracts of 2,421 studies into a corpus, and words with high frequency and high connection centrality were extracted through preprocessing. The analysis revealed a similar word frequency trend between 2010 and 2019 to that between 2020 and 2022. In terms of the research method, many studies using machine learning were conducted from 2010 to 2020, and since 2021, research using deep learning has been increasing. Through these studies, we investigated the trends in research on artificial intelligence utilization by field and the strengths, problems, and challenges of related research. We found that since 2021, the application of artificial intelligence has been expanding, such as research using artificial intelligence for drug rearrangement, using computers to develop anticancer drugs, and applying artificial intelligence to clinical trials. This article briefly presents the prospects of new drug development research using artificial intelligence. If the reliability and safety of bio and medical data are ensured, and the development of the above artificial intelligence technology continues, it is judged that the direction of new drug development using artificial intelligence will proceed to personalized medicine and precision medicine, so we encourage efforts in that field.