• The Journal of the Convergence on Culture Technology
• /
• v.9 no.5
• /
• pp.229-235
• /
• 2023

This study investigates the relationship between syntactic complexity indices and scores on language use in Jacobs et al.(1981)' analytic rating scale. Syntactic complexity indices obtained from TAASSC program and 440 essays written by EFL students from the ICNALE corpus were analyzed. Specifically, this study explores the relationship between scores on language use and Lu(2011)'s traditional syntactic complexity indices, phrasal complexity indices, and clausal complexity indices, respectively. Results of the stepwise regression analysis showed that phrasal complexity indices turned out to be the best predictor of scores on language use, although the variance in scores on language use was relatively small, compared with the previous study. Implications of the findings of the current study for writing instruction (i.e., syntactic structures at the phrase level) were also discussed.

• Journal of Environmental Health Sciences
• /
• v.47 no.4
• /
• pp.310-319
• /
• 2021

Background: Humidifier disinfectant exposure is an ongoing issue, and there is still considerable related controversy. Various approaches are needed to secure scientific evidence on the extent of the victims' damages and for the determination of appropriate compensation. Objectives: The purpose of this study was to assess the association between humidifier disinfectant (HD) use and academic achievement in Korean children. Methods: This study used data from the 8th Panel Study on Korean Children in 2015. For the final study, 1,598 cases were used. T-tests and multiple linear regression analyses were conducted to determine whether the use of humidifier disinfectant is a factor that affects academic ability. Results: Children in groups using humidifier disinfectant showed statistically significantly lower scores in all areas of language, including reading, speaking and writing, and statistically lower scores in all areas of mathematics, including counting, addition and subtraction. In the multiple regression analysis results, which control for the effects of various demographic/social variables, the use of humidifier disinfectants showed statistically significant beta coefficients (β: -0.357, p<0.001), negatively affecting children's language ability. As for the 'math' variable, which was created by combining counting, addition, and subtraction scores, the use of humidifier disinfectants as independent variables also showed statistically significant beta coefficients (β: -0.200, p<0.001), negatively affecting children's math ability. Conclusions: The results of the study showed that depending on whether or not humidifier disinfectants were used, there are differences in children's language abilities, such as reading, speaking, and writing, as well as in their mathematical abilities, such as counting, adding, and subtracting numbers. These findings are thought to serve as a scientific basis for extending the perspective from health effects to more diverse areas of demographic and social impact related to humidifier disinfectant damage and compensation.

• Proceedings of the KSPS conference
• /
• 2004.05a
• /
• pp.239-242
• /
• 2004

An automatic pronunciation correction system provides users with correction guidelines for each pronunciation error. For this purpose, we propose a speech recognition system which automatically classifies pronunciation errors when Koreans speak a foreign language. In this paper, we also propose machine scoring methods for automatic assessment of pronunciation quality by the speech recognizer. Scores obtained from an expert human listener are used as the reference to evaluate the different machine scores and to provide targets when training some of algorithms. We use a log-likelihood score and a normalized log-likelihood score as machine scoring methods. Experimental results show that the normalized log-likelihood score had higher correlation with human scores than that obtained using the log-likelihood score.

• English Language & Literature Teaching
• /
• v.17 no.3
• /
• pp.309-326
• /
• 2011

This study examines the relationship between second language use outside of class and oral proficiency development. It first identifies out-of-class activities of international graduate students in the U.S. and the average time spent speaking English in those out-of-class activities. Interviews and student self-measurements of time spent speaking English each day were used to investigate the types and quantities of out-of-class activities. In addition, two sets of student oral proficiency test scores were collected. Correlation analysis is used to find out the relationship of the variables between the most salient out-of-class activities and oral proficiency gains. The findings indicate that second language use outside of class is important for international graduate students to improve their oral proficiency. This is especially true with regularized interaction such as talking at work and the average time spent speaking in English a day outside of class. This study suggests that learners of English in an ESL environment should be encouraged to take part in out-of-class activities in addition to English use in the classroom in order for them to improve their oral proficiency.

• English Language & Literature Teaching
• /
• v.18 no.2
• /
• pp.189-211
• /
• 2012

The purpose of the study was to explore ways to improve the test items on the Korean scholastic ability test. More specifically, the researchers investigated whether use of the target language in test items would make a difference in total scores, discriminatory power, and item difficulty. A total of 288 high school seniors participated in the study. The subjects were divided into the experimental group (N=145) and the control group (N=143). A 25-item test resembling the Korean scholastic ability test was administered to both groups. The experimental group was given items whose questions and alternatives were all presented in English, whereas the control group was given items whose questions and alternatives were presented in Korean only. Statistical analyses revealed that use of English vs. Korean in the questions and alternatives made a significant difference in total scores, item discrimination, and item difficulty level. The findings strongly suggest that use of English is one way to improve the quality of the Korean scholastic ability test by enhancing item discrimination and face validity. Considering that the test in question is a high-stakes exam in Korea, further research on how to improve the Korean scholastic ability test is urgently called for.

• Asia Pacific Journal of Corpus Research
• /
• v.1 no.2
• /
• pp.1-14
• /
• 2020

Previous work has shown that men both explain and value the act of explaining more than women, as explaining conveys expertise. However, previous studies are limited to English. We conducted an exploratory study to see if similar patterns are seen amongst Japanese speakers. We examined three registers of Japanese: conversational interviews, simulated speeches, and academic presentations. For each text, we calculated two measures: lexical density and the percentage of the text written in kanji. Both are indicators of expository language. Men produced significantly higher scores for the interviews and speeches. However, the results for the presentations depend on age and academic field. In fields in which women are the minority, women produce higher scores. In the field in which men are the minority, younger men produced higher scores but older men produced lower scores than women of the same age. Our results show that in academic contexts, the explainers are not necessarily men but rather the gender minority. We argue that such speakers are under social pressure to present themselves as experts. These results show that the generalization that men tend to explain more than women does not always hold true, and we urge more academic work on expository language.

• Speech Sciences
• /
• v.11 no.2
• /
• pp.217-226
• /
• 2004

An automatic pronunciation correction system provides users with correction guidelines for each pronunciation error. For this purpose, we develop a speech recognition system which automatically classifies pronunciation errors when Koreans speak a foreign language. In this paper, we propose a machine scoring method for automatic assessment of pronunciation quality by the speech recognizer. Scores obtained from an expert human listener are used as the reference to evaluate the different machine scores and to provide targets when training some of algorithms. We use a log-likelihood score and a normalized log-likelihood score as machine scoring methods. Experimental results show that the normalized log-likelihood score had higher correlation with human scores than that obtained using the log-likelihood score.

• English Language & Literature Teaching
• /
• v.10 no.3
• /
• pp.191-210
• /
• 2004

The scores of TOEFL, TOEIC and TEPS have been increasingly used for many purposes in Korea In particular, these test scores are being used as a measure of the test-takers' overall communicative competence. Nonetheless, few studies have been conducted that investigate the validity of the test scores used for this purpose. As a preliminary step to explore the validity of the test scores, we conducted a small scale study by comparing 30 university students' scores in TOEFL, TOElC and TEPS with their class performances in English conversation and composition courses. The correlations between the test scores and the grade point averages of the courses show that the test scores are much harder to use as a valid measure of test-takers' overall communicative competence than usually thought to be and that the score in TOEFL is, nonetheless, a more reliable measure than the ones in the other tests. Although tins study has a few limitations such as the small number of participants, the homogeneity of the sample as a group, etc, it provides some insight into the use of the three tests for measurement of overall communicative proficiency and suggests need for conducting further validation studies in these areas.

• ETRI Journal
• /
• v.46 no.1
• /
• pp.59-70
• /
• 2024

We introduce a high-performance named entity recognition (NER) model for written and spoken language. To overcome challenges related to labeled data scarcity and domain shifts, we use transfer learning to leverage our previously developed KorBERT as the base model. We also adopt a meta-pseudo-label method using a teacher/student framework with labeled and unlabeled data. Our model presents two modifications. First, the student model is updated with an average loss from both human- and pseudo-labeled data. Second, the influence of noisy pseudo-labeled data is mitigated by considering feedback scores and updating the teacher model only when below a threshold (0.0005). We achieve the target NER performance in the spoken language domain and improve that in the written language domain by proposing a straightforward rollback method that reverts to the best model based on scarce human-labeled data. Further improvement is achieved by adjusting the label vector weights in the named entity dictionary.

• ETRI Journal
• /
• v.44 no.5
• /
• pp.794-804
• /
• 2022

Transformer models have had a great impact on natural language processing (NLP) in recent years by realizing outstanding and efficient contextualized language models. Recent studies have used transformer-based language models for various NLP tasks, including Persian named entity recognition (NER). However, in complex tasks, for example, NER, it is difficult to determine which contextualized embedding will produce the best representation for the tasks. Considering the lack of comparative studies to investigate the use of different contextualized pretrained models with sequence modeling classifiers, we conducted a comparative study about using different classifiers and embedding models. In this paper, we use different transformer-based language models tuned with different classifiers, and we evaluate these models on the Persian NER task. We perform a comparative analysis to assess the impact of text representation and text classification methods on Persian NER performance. We train and evaluate the models on three different Persian NER datasets, that is, MoNa, Peyma, and Arman. Experimental results demonstrate that XLM-R with a linear layer and conditional random field (CRF) layer exhibited the best performance. This model achieved phrase-based F-measures of 70.04, 86.37, and 79.25 and word-based F scores of 78, 84.02, and 89.73 on the MoNa, Peyma, and Arman datasets, respectively. These results represent state-of-the-art performance on the Persian NER task.