• Journal of Information Processing Systems
• /
• 제13권4호
• /
• pp.716-730
• /
• 2017

In this paper, we propose a transliteration approach based on semantic information (i.e., language origin and gender) which are automatically learnt from the person name, aiming to transliterate the person name of Uyghur into Chinese. The proposed approach integrates semantic scores (i.e., performance on language origin and gender detection) with general transliteration model and generates the semantic knowledge-based model which can produce the best candidate transliteration results. In the experiment, we use the datasets which contain the person names of different language origins: Uyghur and Chinese. The results show that the proposed semantic transliteration model substantially outperforms the general transliteration model and greatly improves the mean reciprocal rank (MRR) performance on two datasets, as well as aids in developing more efficient transliteration for named entities.

• 영어어문교육
• /
• 제12권1호
• /
• pp.257-286
• /
• 2006

This paper investigates the effects of small group collaborative transcription activities on the improvement of university students' English listening ability. Based on the results of pre-tests, a class of 42 students was subdivided into 11 small homogeneous groups of three to four students. Each group was provided with 7 different news items to be transcribed. Means were compared between pre-test and post-test scores, followed by comparison of transcription errors in content and function words. Finally, students' perception and attitudes about the use of small group transcription for improving listening skills were qualitatively analysed and compared with pre- and post test results. It was found that post test scores highly correlated with improvement of listening ability. It was also found that groups of high degree of collaboration showed improvement in listening ability while lack of collaboration contributed to loss of interests in listening. Students who see process more important were found to make more contribution to the group success than those who place more emphasis on the results.

• 대한음성학회:학술대회논문집
• /
• 대한음성학회 2003년도 10월 학술대회지
• /
• pp.37-40
• /
• 2003

This paper presents the preliminary result of the automatic pronunciation scoring for non-native English speakers, and shows the developmental process for an English speech recognizer for the educational and evaluational purposes. The proposed speech recognizer, featuring two refined acoustic model sets, implements the noise-robust data compensation, phonetic alignment, highly reliable rejection, key-word and phrase detection, easy-to-use language modeling toolkit, etc., The developed speech recognizer achieves 0.725 as the average correlation between the human raters and the machine scores, based on the speech database YOUTH for training and K-SEC for test.

• 정보관리학회지
• /
• 제31권3호
• /
• pp.331-351
• /
• 2014

이 연구는 OECD PISA에 참여한 우리나라 학생, 학부모와 학교장의 설문자료와 읽기 수학 과학소양의 성취도를 분석하여 학교도서관변인이 학업성취도에 미치는 영향력을 파악하고자 하였다. 이를 통해 학교도서관의 교육적 영향력과 책무성에 대한 기초자료를 얻고 학교도서관 교육을 개선하기 위한 시사점을 도출하고자 하였다. 분석결과, 첫째, 학생들이 독서를 즐겁게 생각할수록 학생들의 학업성취도가 높았다. 둘째, 가정에 책이 많고 이용할 수 있는 독서자원이 많을수록 학생들의 학업성취도가 높았다. 셋째, 인터넷을 사용하기 위해 학교도서관을 방문할수록 학업성취도가 낮았다. 넷째, 학교도서관 직원이 부족할수록 학생들의 학업성취도가 낮았다. 즉 학교도서관 직원이 학생들의 학업성취도에 긍정적인 영향을 미치고 있었다. 다섯째, 독서 사교육에 대한 경험은 읽기 과학소양성취도에 부정적인 영향을 미치고 있었다. 여섯째, 독서가 좋아서 학교도서관에서 대출하는 것과 국어교과 관련 학교도서관 수업은 학업성취도에 영향을 미치지 않는 것으로 나타났다.

• International Journal of Internet, Broadcasting and Communication
• /
• 제15권2호
• /
• pp.218-226
• /
• 2023

In Korea, we still use function point based cost estimations for software size and cost of a project. The current problem is that we make difficultly calculating function points with requirements and also have less accurate. That is, it is difficult for non-experts to analyze requirements and calculate function point values with them, and even experts often derive different function points. In addition, all stakeholders strongly make the validity and accuracy of the function point values of the project before /after the development is completed. There are methods for performing function point analysis using source code [1][2][3][4] and some researchers [5][6][7] attempt empirical verification of function points about the estimated cost. There is no research on automatic cost validation with source code after the final development is completed. In this paper, we propose automatically how to calculate Function Points based on natural language requirements before development and prove FP calculation based on the final source code after development. We expect validation by comparing the function scores calculated by forward engineering and reverse engineering methods.

• International Journal of Computer Science & Network Security
• /
• 제24권2호
• /
• pp.143-149
• /
• 2024

The role of social networking mobile applications such as WhatsApp in enhancing second language vocabulary learning among English language learners continues to be a subject of interest for many scholars. The current study aimed at examining medical English vocabulary learning among undergraduate students using WhatsApp compared to learning vocabulary via the Blackboard platform during the Covid-19 pandemic. To this end, 108 medical students (51 males, 57 females) enrolled in a first semester English for a specific English course participated in the study. A quasi-experimental design was adopted for two groups. Fifty-three students participated in the WhatsApp group and 55 students formed the Blackboard group. A pretest-posttest design was employed to collect data. Results of t-test scores did not show a significant difference between the WhatsApp and Blackboard groups on a vocabulary test. Results of a survey that measured students' opinion of the use of WhatsApp as a platform for learning new vocabulary showed positive perceptions since participants thought that WhatsApp enhanced their learning experience.

• 영어어문교육
• /
• 제15권4호
• /
• pp.345-364
• /
• 2009

This study attempts to provide a comprehensive framework for listening strategy use among university students in Korea in relation to TOEIC listening scores. In particular, this study tests whether motivation, based on the self-determination theory, mediates listening strategy use on listening comprehension (LC) process and whether reading comprehension ability moderates the use of listening strategy in LC achievement. One hundred seventy six freshmen students participated in the study during their first semester required English course. Self-report questionnaires were used to assess students' motivation and LC strategy use. The responses were statistically analyzed via the moderator and mediator model. The results indicate that internalized motivation mediates the use of listening strategy in LC achievement; however, reading comprehension skill doesn't affect students' use of listening strategies in relation to listening skill achievement. In other words, students who have internalized motivation were able to utilize listening strategies effectively in terms of achievement of the TOEIC listening skills. The findings of the current study offer in-depth understanding of the relationship among use of LC strategies, intrinsic motivation, and listening skill achievement shared by the mediator and moderator models.

• 영어어문교육
• /
• 제11권3호
• /
• pp.39-56
• /
• 2005

This study is a small-scale replication study of Kubota (1998). Kubota discovered that L1 rhetoric was not the main cause of L2 writing problems. The conclusion of Kubota's study was in contrast to that of contrastive rhetoric research that claimed that the use of L1 rhetorical styles in L2 writing negatively affects the quality of L2 writing, in particular, at the organizational level. According to Kubota(1998), the use of Japanese rhetorical styles did not deteriorate the quality of English writing. By investigating Korean ESL learners, this replication study examined whether L1 and L2 writings have the same rhetorical styles and how L1 rhetorical styles affect the L2 writing quality. The study analyzed Korean and English persuasive writings written by 12 Korean ESL learners in terms of organization. The results show that Korean learners of English used different rhetorical styles in their L1 Korean writing and L2 English writing. However, there was a positive relationship between L1 Korean and L2 English organization scores.

• 영어어문교육
• /
• 제17권3호
• /
• pp.57-71
• /
• 2011

The purpose of this study is to investigate the possibility of the use of reading aloud as a way of teaching English reading in a formal college-level English reading classroom. The study tried to examine the effects of reading aloud on the development of English reading ability in terms of comprehension and speed and the students' perceptions of the use of reading aloud in an English reading classroom. The participants of the study were 36 third year students at a college. The results of the cloze test and reading speed test showed that reading aloud had a positive effect on the improvement of English reading ability. The students received higher scores in the cloze test after the course and read a text faster than before the course. The analysis of the course evaluation questionnaire indicated that the students in the study showed positive attitudes toward the use of reading aloud in an English reading classroom. Most of the students reported that reading aloud was an effective way of improving their English reading comprehension and speed. However, the students were not fully satisfied with the reading materials used in the study. The results of the present study indicated that reading aloud may be incorporated into a formal English reading classroom effectively.

• Communications for Statistical Applications and Methods
• /
• 제27권3호
• /
• pp.313-326
• /
• 2020

Deep Learning is the most important key to the development of Artificial Intelligence (AI). There are several distinguishable architectures of neural networks such as MLP, CNN, and RNN. Among them, we try to understand one of the main architectures called Recurrent Neural Network (RNN) that differs from other networks in handling sequential data, including time series and texts. As one of the main tasks recently in Natural Language Processing (NLP), we consider Neural Machine Translation (NMT) using RNNs. We also summarize fundamental structures of the recurrent networks, and some topics of representing natural words to reasonable numeric vectors. We organize topics to understand estimation procedures from representing input source sequences to predict target translated sequences. In addition, we apply multiple translation models with Gated Recurrent Unites (GRUs) in Keras on English-Korean sentences that contain about 26,000 pairwise sequences in total from two different corpora, colloquialism and news. We verified some crucial factors that influence the quality of training. We found that loss decreases with more recurrent dimensions and using bidirectional RNN in the encoder when dealing with short sequences. We also computed BLEU scores which are the main measures of the translation performance, and compared them with the score from Google Translate using the same test sentences. We sum up some difficulties when training a proper translation model as well as dealing with Korean language. The use of Keras in Python for overall tasks from processing raw texts to evaluating the translation model also allows us to include some useful functions and vocabulary libraries as well.