• Journal of Information Technology Services
• /
• v.21 no.2
• /
• pp.85-95
• /
• 2022

This study analyzed the translation error factors of machine translation services such as Naver Papago and Google Translate through Self-Attention path visualization. Self-Attention is a key method of the Transformer and BERT NLP models and recently widely used in machine translation. We propose a method to explain translation error factors of machine translation algorithms by comparison the Self-Attention paths between ST(source text) and ST'(transformed ST) of which meaning is not changed, but the translation output is more accurate. Through this method, it is possible to gain explainability to analyze a machine translation algorithm's inside process, which is invisible like a black box. In our experiment, it was possible to explore the factors that caused translation errors by analyzing the difference in key word's attention path. The study used the XLM-RoBERTa multilingual NLP model provided by exBERT for Self-Attention visualization, and it was applied to two examples of Korean-Chinese and Korean-English translations.

• The Journal of Korean Academic Society of Nursing Education
• /
• v.18 no.2
• /
• pp.197-205
• /
• 2012

Purpose: This study aimed to demonstrate correlations among self-regulated learning, attention control and Yangseng, to clarify any differences depending on general characteristics and ultimately to understand factors affecting self-regulated learning of undergraduates. Methods: Data were collected for a month from April 1st, 2011. A total of 438 undergraduate nursing students of two universities in Chungbuk and Chungnam were surveyed by using a questionnaire about self-regulated learning, attention control and Yangseng. Results: Self-regulated learning of the subjects showed statistically significant correlations with their attention control (r=.302, p=.001) and Yangseng (r=.292, p=.001). In addition, self-regulated learning could be explained by attention control (${\beta}$=3.648, p=.001), Yangseng (${\beta}$=3.645, p=.001), perceived academic achievement levels (${\beta}$=.124, p=.018), or eating breakfast (${\beta}$=.102, p=.027). In the model, the variables explained self-regulated learning by 19.0%. Conclusion: Nursing instructors should encourage undergraduate nursing students to enhance their attention control so that they can improve their self-regulated learning abilities, which will eventually develop their problem solving skills. In addition, it was shown that self-regulated learning correlates with yangseng including eating a regular breakfast. Maintaining a desirable lifestyle is also essential for students to succeed in self-regulated learning.

• Journal of IKEEE
• /
• v.27 no.3
• /
• pp.251-257
• /
• 2023

In this paper, we propose a dual-structured self-attention method that improves the lack of regional features of the vision transformer's self-attention. Vision Transformers, which are more computationally efficient than convolutional neural networks in object classification, object segmentation, and video image recognition, lack the ability to extract regional features relatively. To solve this problem, many studies are conducted based on Windows or Shift Windows, but these methods weaken the advantages of self-attention-based transformers by increasing computational complexity using multiple levels of encoders. This paper proposes a dual-structure self-attention using self-attention and neighborhood network to improve locality inductive bias compared to the existing method. The neighborhood network for extracting local context information provides a much simpler computational complexity than the window structure. CIFAR-10 and CIFAR-100 were used to compare the performance of the proposed dual-structure self-attention transformer and the existing transformer, and the experiment showed improvements of 0.63% and 1.57% in Top-1 accuracy, respectively.

• Journal of Advanced Navigation Technology
• /
• v.27 no.2
• /
• pp.182-189
• /
• 2023

Depth estimation is a key technology in 3D map generation for autonomous driving of vehicles, robots, and drones. The existing sensor-based method has high accuracy but is expensive and has low resolution, while the camera-based method is more affordable with higher resolution. In this study, we propose self-attention-based unsupervised monocular depth estimation for UAV camera system. Self-Attention operation is applied to the network to improve the global feature extraction performance. In addition, we reduce the weight size of the self-attention operation for a low computational amount. The estimated depth and camera pose are transformed into point cloud. The point cloud is mapped into 3D map using the occupancy grid of Octree structure. The proposed network is evaluated using synthesized images and depth sequences from the Mid-Air dataset. Our network demonstrates a 7.69% reduction in error compared to prior studies.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.3
• /
• pp.721-739
• /
• 2023

Rating prediction is an important issue in recommender systems, and its accuracy affects the experience of the user and the revenue of the company. Traditional recommender systems use Factorization Machinesfor rating predictions and each feature is selected with the same weight. Thus, there are problems with inaccurate ratings and limited data representation. This study proposes a deep recommendation model based on self-attention Factorization (SAFMR) to solve these problems. This model uses Convolutional Neural Networks to extract features from user and item reviews. The obtained features are fed into self-attention mechanism Factorization Machines, where the self-attention network automatically learns the dependencies of the features and distinguishes the weights of the different features, thereby reducing the prediction error. The model was experimentally evaluated using six classes of dataset. We compared MSE, NDCG and time for several real datasets. The experiment demonstrated that the SAFMR model achieved excellent rating prediction results and recommendation correlations, thereby verifying the effectiveness of the model.

• Annual Conference on Human and Language Technology
• /
• 2019.10a
• /
• pp.362-367
• /
• 2019

본 논문에서는 Self-Attention을 활용한 딥러닝 기반 문맥의존 철자오류 교정 모델을 제안한다. 문맥의존 철자오류 교정은 최근 철자오류 교정 분야에서 활발히 연구되고 있는 문제 중 하나이다. 기존에는 규칙 기반, 확률 기반, 임베딩을 활용한 철자오류 교정이 연구되었으나, 아직 양질의 교정을 수행해내기에는 많은 문제점이 있다. 따라서 본 논문에서는 기존 교정 모델들의 단점을 보완하기 위해 Self-Attention을 활용한 문맥의존 철자오류 교정 모델을 제안한다. 제안 모델은 Self-Attention을 활용하여 기존의 임베딩 정보에 문맥 의존적 정보가 반영된 더 나은 임베딩을 생성하는 역할을 한다. 전체 문장의 정보가 반영된 새로운 임베딩을 활용하여 동적으로 타겟 단어와의 관련 단어들을 찾아 문맥의존 철자 오류교정을 시행한다. 본 논문에서는 성능평가를 위해 세종 말뭉치를 평가 데이터로 이용하여 제안 모델을 실험하였고, 비정형화된 구어체(Kakao Talk) 말뭉치로도 평가 데이터를 구축해 실험한 결과 비교 모델보다 높은 정확율과 재현율의 성능향상을 보였다.

• The Korean Journal of Community Living Science
• /
• v.25 no.3
• /
• pp.303-319
• /
• 2014

This study considers structural equation model and examines the-relationships between various model variables to identify the causal relationships of between the attention span temperament, affectionate rearing attitudes of mothers, and family support (individual external variables) to children's behavioral problems though self-reliance, psychological variable. According to the results, the attention span temperament had significant direct, indirect, and total effects on the externalization of behavioral problems, but affectionate rearing attitudes and family support had only significant indirect effects. Self-resilience had a significant direct effect. These results can be used as basic data to prevent behavioral problems of children and increase their self-resilience.

• Journal of Oriental Neuropsychiatry
• /
• v.23 no.2
• /
• pp.49-58
• /
• 2012

Objectives : The purpose of this study is to examine the effects of Mindfulness-based concentration qigong for children (MBCQ-C) in healthy children with subjective poor attention. Methods : This study examined the effects of MBCQ-C on healthy children with subjective poor attention, who vistied Korean medicine hospital neuropsychiatry outpatient clinic. The MBCQ-C was practiced with 11 participants, 2 of them quit in the middle of the program, and hence, they were excluded for data analysis. MBCQ-C consisted of 8 sessions, and each session took about 60 minutes. The outcome measurement was Frankfurter Aufmerksamkeits-Inventar (FAIR), which measured selective attention, self-control and sustained attention. Results : The results of this study showed that selective attention, and sustained attention were significantly improved. Self-control also improved, but without any statistical significance. These results indicate MBCQ-C was effective for the improvement of attention abilities, but self-control, including upper cognition area needs more consistent exercise. Conclusions : The MBCQ-C consisting of 8 sessions were shown to be an effective intervention in improving the attention abilities of healthy children with subjective poor attention.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.18 no.11
• /
• pp.628-636
• /
• 2017

This study was conducted to analyze the effects of parents' self-esteem on preschool aged children's emotional responses and attention, mediated by their self-esteem, utilizing data from the 7th Panel Study of Korean Children. This study analyzed the main survey of the 7th year survey of the Panel Study on Korean Children and 1383 families from which both parents participated in the survey (1383 couples of parents and 1383 children). The results revealed that mother's self-esteem had a direct effect on children's self-esteem and that children's self-esteem had direct effects on their emotional responses and attention. Mother's self-esteem also had direct effects on children's emotional responses and attention, as well as indirect effects on their emotional responsiveness and attention, and these effects were mediated by parents' self-esteem. Overall, the study revealed the impact of parents' self-esteem on children's emotional responsiveness and attention and provided basic data for the development of an education program for preschool aged children and parents. Thus, it is necessary to develop educational programs to improve preschool aged children's self-esteem and to develop a program for the formation and maintenance of mother's positive self-esteem.

• Journal of the Korean Academy of Child and Adolescent Psychiatry
• /
• v.28 no.4
• /
• pp.252-259
• /
• 2017

Objective: The purpose of this study was to investigate the impact of clinical and psychological factors on the self-concept of children and adolescents with epilepsy. Methods: Children and adolescents with epilepsy (n=60; age range=9-17 years) completed questionnaires about their epilepsy-related variables, self-concept, depressive symptoms, anxiety, family functions, and behavioral problems. The T-test and one-way analysis of variance were used to examine the variables affecting the total self-concept scores. To determine the independent variables by adjusting the significant variables, a stepwise regression analysis was performed. Results: In the correlational analysis, age, depressive symptoms, anxiety, social problems, attention problems, and internalizing problems had significantly negative correlations with self-concept. On the other hand, IQ and family functions showed positive correlations with selfconcept. Age (${\beta}=-0.177$, p=0.015), depressive symptoms (${\beta}=-0.487$, p<0.001), anxiety (${\beta}=-0.298$, p=0.008), and attention problems (${\beta}=-0.138$, p=0.048) were analyzed as independent factors to assess their impact on self-concept, and were found to account for 78.3% of the variance in self-concept by stepwise regression analysis. Conclusion: Parents and clinicians should pay attention to improving the self-concept of children and adolescents with epilepsy, especially if they have problems with depression, anxiety, or attention.