• ETRI Journal
• /
• v.46 no.1
• /
• pp.48-58
• /
• 2024

This paper presents the development of language tutoring systems for nonnative speakers by leveraging advanced end-to-end automatic speech recognition (ASR) and proficiency evaluation. Given the frequent errors in non-native speech, high-performance spontaneous speech recognition must be applied. Our systems accurately evaluate pronunciation and speaking fluency and provide feedback on errors by relying on precise transcriptions. End-to-end ASR is implemented and enhanced by using diverse non-native speaker speech data for model training. For performance enhancement, we combine semisupervised and transfer learning techniques using labeled and unlabeled speech data. Automatic proficiency evaluation is performed by a model trained to maximize the statistical correlation between the fluency score manually determined by a human expert and a calculated fluency score. We developed an English tutoring system for Korean elementary students called EBS AI Peng-Talk and a Korean tutoring system for foreigners called KSI Korean AI Tutor. Both systems were deployed by South Korean government agencies.

• Journal of Fisheries and Marine Sciences Education
• /
• v.28 no.3
• /
• pp.808-817
• /
• 2016

This study aims at reviewing differences of awareness toward speech-language therapy according to background variables of child-care teachers, and establishing a basic data necessary for special education support programs afterwards based on child-care centers and its characteristics with the results. Researcher carried out a survey by objecting child-care teachers of Chungcheongbuk-do, Korea, and looked into existence of some differences through SPSS 20.0 for Window, independent sample t-Test, ANOVA, Scheffe post verification on background characteristics such as level of education, working career, license grade, whether or not to have objects of speech-language objects in the class. Research results are same as follows. First, significant differences in awareness of child-care teachers toward speech-language therapy appeared from license grade among characteristics like level of education, license grade, whether or not to have objects of speech-language therapy in the class. Second, significant differences were displayed from whether or not to have objects of speech-language therapy in the class among characteristics such as level of education, working career, license grade, and whether or not to have objects of speech-language therapy. When putting these results together, a conclusion could be made such like awareness of child-care teachers toward speech-language therapy and therapists is able to be different a little according to background variables of teachers.

• Smart Media Journal
• /
• v.1 no.3
• /
• pp.8-14
• /
• 2012

In this paper, we develope the smart phone App. contents of 3D sign language to widen the opportunity of the korean sign language education for the hearing-impaired and normal people. Especially, we propose the sign language conversion algorithm that automatically transform the structure of Korean phrases to the structure of the sign language. Also, we implement the 3D sign language animation DB using motion capture system and data glove for acquiring the natural motions. Finally, UNITY 3D engine is used for the realtime 3D rendering of sign language motion. We are distributing the proposed App. with 3D sign language DB of 1,300 words to the iPhone App. store and Android App. store.

• International Journal of Internet, Broadcasting and Communication
• /
• v.14 no.4
• /
• pp.189-197
• /
• 2022

Today, we are exposed to various text-based media such as newspapers, Internet articles, and SNS, and the amount of text data we encounter has increased exponentially due to the recent availability of Internet access using mobile devices such as smartphones. Collecting useful information from a lot of text information is called text analysis, and in order to extract information, it is performed using technologies such as Natural Language Processing (NLP) for processing natural language with the recent development of artificial intelligence. For this purpose, a morpheme analyzer based on everyday language has been disclosed and is being used. Pre-learning language models, which can acquire natural language knowledge through unsupervised learning based on large numbers of corpus, are a very common factor in natural language processing recently, but conventional morpheme analysts are limited in their use in specialized fields. In this paper, as a preliminary work to develop a natural language analysis language model specialized in the railway field, the procedure for construction a corpus specialized in the railway field is presented.

• Journal of Christian Education in Korea
• /
• v.61
• /
• pp.165-201
• /
• 2020

The purpose of this study is to explore the development of language education of Christian early childhood education by exploring the change of pre-Christian preschool teachers through the development of biblical language education activities. Interviews, surveys, action plans, and reflections of 19 Christian education students who participated in the development of bible-based language education activities for children based on the language of early childhood language, were conducted from September 3 to December 28, 2018. The data were collected through a portfolio. By analyzing the collected data, the key categories were derived and categorized. For the objectification of data analysis and interpretation, two thematic and early childhood education specialists were identified. As a result, the preparatory Christian teacher experience for the development of bible-based langage education activities for young children was categorized into cognitive change, personality change and practical change. First, through the development of bible-based early childhood language education activities, Christian pre-service early childhood teachers have brought cognitive changes as 'processes not outcomes', 'integration not separation', 'living non-curriculum' and 'meaning not effect'. In developing bible-based language education activities for young children, Christian pre-service early childhood teachers experienced a cognitive change in the 'process of language education activities' rather than the developmental achievements and results of early childhood language education. Christian pre-service early childhood teachers recognized the necessity of 'integration of listening-speaking-reading-writing', not the separation of early childhood language education. They recognized the importance of 'informal language education in kindergarten life', as well as teacher-centered formal language education. In addition, they have made a cognitive change that 'child-centered meaningful language education experience' is more important than the effectiveness of early childhood language education. Second, in developing bible-based language education activities for young children, Christian pre-service early childhood teachers showed personality changes as 'confident teachers', 'professional teachers', and 'teachers with reflective thoughts and attitudes'. Finally, in developing bible-based language education activities for young children, Christian pre-service early childhood teachers recognized the power of positive language and practiced it to form habits of using the right language and to link Christian education with early childhood education. Through the development of bible-based language education activities for young children, Christian pre-service early childhood teachers are equipped with the heart attitude and enthusiasm required to become true early childhood teachers for young children in unpredictable educational conditions and rapidly changing educational realities. Teacher efficacy has improved. In the future, it is expected that various teacher education programs linking Christian education and early childhood education will be continuously and systematically implemented.

• Journal of The Korean Society of Integrative Medicine
• /
• v.7 no.1
• /
• pp.37-45
• /
• 2019

Purpose : The aim of this study was to investigate the differences in sociodemographic characteristics, instrumental activities of daily living (IADL), and healthcare needs among disabled persons with and without language disorder. Methods : Using raw data from the National Survey of the Disabled Person (2017), this study compared 6,320 disabled persons without language disorder and 229 disabled person with language disorder among 6,549 persons with disabilities. The dependent variables included sociodemographic characteristics, IADL, and healthcare needs. Results : People with language disorder were younger (p<.001), illiterate (p<.001), and had higher family income (p=.003) than people without language disorder. In addition, people with language disorder had a higher disability grade than people without language disorder (p<.001), and the percentages of brain injury, hearing impairment, and intellectual disability were higher among type of disability (p<.001). People with language disorder showed significantly higher dependence on all items of IADL than people without language disorder (p<.001). As the first-ranked healthcare need, people without language disorder had the highest percentage of "chronic illness management," and people with language disorder had the highest percentage of "disability management" (p<.001). Conclusion : The findings of the current investigation suggest that health care services are needed to improve IADL in people with language disorder and that the development of a disability management program for people with language disorder is required.

• Proceedings of the Korean Society for Language and Information Conference
• /
• 1998.02a
• /
• pp.295-302
• /
• 1998

• Journal of the Korea Society of Computer and Information
• /
• v.27 no.11
• /
• pp.19-27
• /
• 2022

In this study, we propose a language and platform to describe and manage the MLOps(Machine Learning Operations) workflow for time series data anomaly detection. Time series data is collected in many fields, such as IoT sensors, system performance indicators, and user access. In addition, it is used in many applications such as system monitoring and anomaly detection. In order to perform prediction and anomaly detection of time series data, the MLOps platform that can quickly and flexibly apply the analyzed model to the production environment is required. Thus, we developed Python-based AI/ML Modeling Language (AMML) to easily configure and execute MLOps workflows. Python is widely used in data analysis. The proposed MLOps platform can extract and preprocess time series data from various data sources (R-DB, NoSql DB, Log File, etc.) using AMML and predict it through a deep learning model. To verify the applicability of AMML, the workflow for generating a transformer oil temperature prediction deep learning model was configured with AMML and it was confirmed that the training was performed normally.

• Proceedings of the CALSEC Conference
• /
• 1999.11a
• /
• pp.75-84
• /
• 1999

CALS is a strategy to share integrated product data through a set of standards to achieve efficiencies in business and operational mission areas. In this research, we first studied current status for CALS standard and then analyzed the case of US DoD. The results can be summarized as implementing for two major standard in defense CALS system. They are STEP and XML. Our recommendations can be used to set direction for CALS standard implementation and standard selection process.

• 한국HCI학회:학술대회논문집
• /
• 2008.02a
• /
• pp.384-389
• /
• 2008

Recently, the necessity of systems which effectively store and retrieve video data has increased. Conventional video retrieval systems retrieve data using menus or text based keywords. Due to the lack of information, many video clips are simultaneously searched, and the user must have a certain level of knowledge to utilize the system. In this paper, we suggest a natural language based conversational video retrieval system that reflects users' intentions and includes more information than keyword based queries. This system can also retrieve from events or people to their movements. First, an event database is constructed based on meta-data which are generated by domain analysis for collected video in an office environment. Then, a script database is also constructed based on the query pre-processing and analysis. From that, a method to retrieve a video through a matching technique between natural language queries and answers is suggested and validated through performance and process evaluation for 10 users The natural language based retrieval system has shown its better efficiency in performance and user satisfaction than the menu based retrieval system.