• Annual Conference on Human and Language Technology
• /
• 2021.10a
• /
• pp.340-345
• /
• 2021

최근 많은 연구들이 BERT를 활용하여, 주어진 문맥에서 언어학/문법적으로 적절하지 않은 단어를 인지하고 찾아내는 성과를 보고하였다. 하지만 일반적으로 딥러닝 관점에서 NLL기법(Negative log-likelihood)은 주어진 문맥에서 언어 변칙에 대한 정확한 성격을 규명하기에는 어려움이 있다고 지적되고 있다. 이러한 한계를 해결하기 위하여, Li et al.(2021)은 트랜스포머 언어모델의 은닉층별 밀도 추정(density estimation)을 통한 가우시안 확률 분포를 활용하는 가우시안 혼합 모델(Gaussian Mixture Model)을 적용하였다. 그들은 트랜스포머 언어모델이 언어 변칙 예문들의 종류에 따라 상이한 메커니즘을 사용하여 처리한다는 점을 보고하였다. 이 선행 연구를 받아들여 본 연구에서는 한국어 기반 언어모델인 KoBERT나 KR-BERT도 과연 한국어의 상이한 유형의 언어 변칙 예문들을 다른 방식으로 처리할 수 있는지를 규명하고자 한다. 이를 위해, 본 연구에서는 한국어 형태통사적 그리고 의미적 변칙 예문들을 구성하였고, 이 예문들을 바탕으로 한국어 기반 모델들의 성능을 놀라움-갭(surprisal gap) 점수를 계산하여 평가하였다. 본 논문에서는 한국어 기반 모델들도 의미적 변칙 예문을 처리할 때보다 형태통사적 변칙 예문을 처리할 때 상대적으로 보다 더 높은 놀라움-갭 점수를 보여주고 있음을 발견하였다. 즉, 상이한 종류의 언어 변칙 예문들을 처리하기 위하여 다른 메커니즘을 활용하고 있음을 보였다.

• Journal of the Korea Society of Computer and Information
• /
• v.24 no.11
• /
• pp.41-49
• /
• 2019

We are to build an unsupervised machine learning-based language model which can estimate the amount of information that are in need to process words consisting of subword-level morphemes and syllables. We are then to investigate whether the reading times of words reflecting their morphemic and syllabic structures are predicted by an information-theoretic measure such as surprisal. Specifically, the proposed Morfessor-based unsupervised machine learning model is first to be trained on the large dataset of sentences on Sejong Corpus and is then to be applied to estimate the information-theoretic measure on each word in the test data of Korean words. The reading times of the words in the test data are to be recruited from Korean Lexicon Project (KLP) Database. A comparison between the information-theoretic measures of the words in point and the corresponding reading times by using a linear mixed effect model reveals a reliable correlation between surprisal and reading time. We conclude that surprisal is positively related to the processing effort (i.e. reading time), confirming the surprisal hypothesis.

• Journal of the Korea Society of Computer and Information
• /
• v.25 no.11
• /
• pp.27-40
• /
• 2020

In this paper, we investigate the correlation between the amount of English sentences that Korean English learners (L2ers) are exposed to and their sentence processing patterns by examining what Long Short-Term Memory (LSTM) language models (LMs) can learn about implicit syntactic relationship: that is, the filler-gap dependency. The filler-gap dependency refers to a relationship between a (wh-)filler, which is a wh-phrase like 'what' or 'who' overtly in clause-peripheral position, and its gap in clause-internal position, which is an invisible, empty syntactic position to be filled by the (wh-)filler for proper interpretation. Here to implement L2ers' English learning, we build LSTM LMs that in turn learn a subset of the known restrictions on the filler-gap dependency from English sentences in the L2 corpus that L2ers can potentially encounter in their English learning. Examining LSTM LMs' behaviors on controlled sentences designed with the filler-gap dependency, we show the characteristics of L2ers' sentence processing using the information-theoretic metric of surprisal that quantifies violations of the filler-gap dependency or wh-licensing interaction effects. Furthermore, comparing L2ers' LMs with native speakers' LM in light of processing the filler-gap dependency, we not only note that in their sentence processing both L2ers' LM and native speakers' LM can track abstract syntactic structures involved in the filler-gap dependency, but also show using linear mixed-effects regression models that there exist significant differences between them in processing such a dependency.

• Journal of the Korean Society of Clothing and Textiles
• /
• v.41 no.2
• /
• pp.266-280
• /
• 2017

This study predicted consumer approach/avoidance behavior through consumer emotional experiences and examined the moderating effect of perceived surprises in the context of digital signage in store environments. A self-administered questionnaire consisted of consumer emotional experience (e.g., pleasure, arousal, and dominance), approach-avoidance behavior and perceived surprise by digital signage. A total of 278 usable responses were obtained from consumers who experienced digital signage at fashion retail stores. The findings support the Mehrabian-Russell model in the context of digital signage. Approach behavior was predicted by pleasure and arousal emotional experience, while avoidance behavior was predicted by dominance. The moderating effect of perceived surprise was also indicated in the effect of emotional experience on approach or avoidance behavior. In the high level of perceived surprise, pleasure and arousal had significant effects on approach behavior, whereas dominance had significant effect on avoidance behavior. This study discussed theoretical and managerial implications for creating emotional experiences and developing strategic store management by utilizing new digital technology within the fashion retail environments.

• Fire Science and Engineering
• /
• v.23 no.2
• /
• pp.67-77
• /
• 2009

The purpose of this study is to deal with fire fighting safety education primarily among the fields of child safety education. So, first of all, this study considered the theoretical background of fire fighting safety education. And this study analyzed the present state and cases of safety accidents in elementary schools. And by focusing on the fire fighting science class for elementary schools which Gyeonggi-do Goyang fire station is executing now, this study analyzed the educational outline, present state, educational goal, and content of fire fighting science class, and conducted the satisfaction survey through questionnaire over the elementary school students participating in fire fighting science class and the fire fighting officers in charge of fire fighting science class. On basis of this research, by developing the new field of fire fighting education and publicity into the program which can diffuse the fire fighting-related chemical experiment based on the science of chemistry and physics in the future and so provide the pleasure and surprise of experiencing directly not only natural fire fighting education and publicity but also learning and the common sense of fire fighting, this study tried to present the alternatives about the measures for activating the fire fighting safety education in elementary schools.

• The Journal of the Korea Contents Association
• /
• v.11 no.12
• /
• pp.182-189
• /
• 2011

This study was conducted to evaluate the usability of augmented reality (AR) picture book for young children. The participants of this study were 17 five-year-olds attending to M kindergarten in Korea. The measurement of the usability was conducted by observation and interviews during and after children's AR picture book reading 1) usability test score such as task completed that is, reading the AR picturebook successfully, 2) children's behaviors in the process of reading the book, 3) children's response to the questions asking the level of satisfaction with the AR picture book. The result of study showed that most of children could read the AR successfully adjusting the distance and angle of AR picture book to web-camera's location. And children showed verbal and nonverbal expression such as surprise, joy, questions and frustration in reading the book. Children told that AR picture book is fun and easy to read. This study would provide the implications for the interface development and adult-child book reading of AR picture books.

• Proceedings of the Korea Contents Association Conference
• /
• 2006.05a
• /
• pp.160-163
• /
• 2006

If an instruction system understood the interest and activeness of a learner in real time, it could provide some interesting factors when a learner is tired of learning. It could work as an adaptive tutoring system to help a learner to understand something difficult to understand. Currently the area of the facial expression recognition mainly deals with the facial expression of adults focusing on anger, hatred, fear, sadness, surprising and gladness. These daily facial expressions couldn't be one of expressions of a learner in e-Learning. They should first study the facial expressions of a learner in e-Learning to recognize the feeling of a learner. Collecting as many expression pictures as possible, they should study the meaning of each expression. This study, as a prior research, analyzes the feelings of learners and facial expressions of learners in e-Learning in relation to the feelings to establish the facial expressions database.

• Journal of The Korean Association of Information Education
• /
• v.26 no.5
• /
• pp.327-340
• /
• 2022

This study applied effective playful learning to increase the learning motivation of non-CS major students to help them achieve learning and to successfully operate online SW liberal arts classes for 560 students. As a result of analyzing the students' reflection journals, most of the students accepted the 'white radish' of dialect names as fun playful learning in the process of learning local variables and global variables. And they were surprised and amazed at discovering unexpected contents in our SW class. It was found that they experienced delight in learning, learning-flow, confidence, and intrinsic motivation. In the final term exam at the end of the semester, it was confirmed that the correct rate of 92% for questions related to local and global variables was higher than the average rate of other questions' correctness of 67.1%.

• Proceedings of the Korea Institute of Fire Science and Engineering Conference
• /
• 2008.11a
• /
• pp.270-276
• /
• 2008

The purpose of this study is to deal with fire fighting safety education primarily among the fields of child safety education. So, first of all, this study considered the theoretical background of fire fighting safety education. And this study analyzed the present state and cases of safety accidents in elementary schools. And by focusing on the fire fighting science class for elementary schools which Gyeonggi-do Goyang fire station is executing now, this study analyzed the educational outline, present state, educational goal, and content of fire fighting science class, and conducted the satisfaction survey through questionnaire over the elementary school students participating in fire fighting science class and the fire fighting officers in charge of fire fighting science class. On basis of this research, by developing the new field of fire fighting education and publicity into the program which can diffuse the fire fighting-related chemical experiment based on the science of chemistry and physics in the future and so provide the pleasure and surprise of experiencing directly not only natural fire fighting education and publicity but also learning and the common sense of fire fighting, this study tried to present the alternatives about the measures for activating the fire fighting safety education in elementary schools.

• Information Systems Review
• /
• v.17 no.2
• /
• pp.133-156
• /
• 2015

A recent interesting user experience in social curation websites is that they often encounter unanticipated, valuable information through accidental, incidental, or serendipitous discoveries. Prior research suggests that serendipity can help solve a problem, generate a creative idea, and invent a new product because it allows for an intellectual leap to arrive at a novel insight that is suddenly instrumental to a task. Despite the significant roles, the notion of serendipity is relatively unexplored especially in the context where people actually experience it, such as social curation websites. In this study, we theoretically identified the key technological features of social curation websites, which support three major curation activities respectively: browsing, collection, and interaction. We then examined their impacts on the occurrence of serendipitous discoveries in the process of information seeking. Our results using data collected from real users reveal that the technological features all affected user experience of serendipity and the interaction feature has stronger impacts than the other two. Moreover, it was also found that user experience of serendipity significantly affected website usage behavior such as intention to return and intention to contribute to the website.