• Journal of the Institute of Convergence Signal Processing
• /
• v.24 no.3
• /
• pp.160-165
• /
• 2023

In this study, we propose an SNS sentence writing assistance system that utilizes YOLO and GPT to assist users in writing texts with images, such as SNS. We utilize the YOLO model to extract objects from images inserted during writing, and also extract meta-information such as GPS information and creation time information, and use them as prompt values for GPT. To use the YOLO model, we trained it on form image data, and the mAP score of the model is about 0.25 on average. GPT was trained on 1,000 blog text data with the topic of 'restaurant reviews', and the model trained in this study was used to generate sentences with two types of keywords extracted from the images. A survey was conducted to evaluate the practicality of the generated sentences, and a closed-ended survey was conducted to clearly analyze the survey results. There were three evaluation items for the questionnaire by providing the inserted image and keyword sentences. The results showed that the keywords in the images generated meaningful sentences. Through this study, we found that the accuracy of image-based sentence generation depends on the relationship between image keywords and GPT learning contents.

• Journal of The Korean Astronomical Society
• /
• v.56 no.2
• /
• pp.213-224
• /
• 2023

Type II solar radio bursts show frequency drifts from high to low over time. They have been known as a signature of coronal shock associated with Coronal Mass Ejections (CMEs) and/or flares, which cause an abrupt change in the space environment near the Earth (space weather). Therefore, early detection of type II bursts is important for forecasting of space weather. In this study, we develop a deep-learning (DL) model for the automatic detection of type II bursts. For this purpose, we adopted a 1-D Convolution Neutral Network (CNN) as it is well-suited for processing spatiotemporal information within the applied data set. We utilized a total of 286 radio burst spectrum images obtained by Hiraiso Radio Spectrograph (HiRAS) from 1991 and 2012, along with 231 spectrum images without the bursts from 2009 to 2015, to recognizes type II bursts. The burst types were labeled manually according to their spectra features in an answer table. Subsequently, we applied the 1-D CNN technique to the spectrum images using two filter windows with different size along time axis. To develop the DL model, we randomly selected 412 spectrum images (80%) for training and validation. The train history shows that both train and validation losses drop rapidly, while train and validation accuracies increased within approximately 100 epoches. For evaluation of the model's performance, we used 105 test images (20%) and employed a contingence table. It is found that false alarm ratio (FAR) and critical success index (CSI) were 0.14 and 0.83, respectively. Furthermore, we confirmed above result by adopting five-fold cross-validation method, in which we re-sampled five groups randomly. The estimated mean FAR and CSI of the five groups were 0.05 and 0.87, respectively. For experimental purposes, we applied our proposed model to 85 HiRAS type II radio bursts listed in the NGDC catalogue from 2009 to 2016 and 184 quiet (no bursts) spectrum images before and after the type II bursts. As a result, our model successfully detected 79 events (93%) of type II events. This results demonstrates, for the first time, that the 1-D CNN algorithm is useful for detecting type II bursts.

• Journal of Digital Policy
• /
• v.2 no.4
• /
• pp.1-10
• /
• 2023

This study conducted preliminary study to identify effective ways to use ChatGPT in traffic policing by analyzing ChatGPT's responses to the driver's license test and the road traffic accident appraiser test. I collected ChatGPT responses for the driver's license test item pool and the road traffic accident appraiser test using the OpenAI API with Python code for 30 iterative experiments, and analyzed the percentage of correct answers by test, year, section, and consistency. First, the average correct answer rate for the driver's license test and the for road traffic accident appraisers test was 44.60% and 35.45%, respectively, which was lower than the pass criteria, and the correct answer rate after 2022 was lower than the average correct answer rate. Second, the percentage of correct answers by section ranged from 29.69% to 56.80%, showing a significant difference. Third, it consistently produced the same response more than 95% of the time when the answer was correct. To effectively utilize ChatGPT, it is necessary to have user expertise, evaluation data and analysis methods, design a quality traffic law corpus and periodic learning.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2024.01a
• /
• pp.55-58
• /
• 2024

최근 각 기업의 AI 면접시스템 도입이 증가하고 있으며, AI 면접에 대한 실효성 논란 또한 많은 상황이다. 본 논문에서는 AI 면접 과정에서 지원자를 평가하는 방식을 시각, 음성, 자연어처리 3영역에서 구현함으로써, 면접 지원자를 다방면으로 분석 방법론의 적절성에 대해 평가하고자 한다. 첫째, 시각적 측면에서, 면접 지원자의 감정을 인식하기 위해, 합성곱 신경망(CNN) 기법을 활용해, 지원자 얼굴에서 6가지 감정을 인식했으며, 지원자가 카메라를 응시하고 있는지를 시계열로 도출하였다. 이를 통해 지원자가 면접에 임하는 태도와 특히 얼굴에서 드러나는 감정을 분석하는 데 주력했다. 둘째, 시각적 효과만으로 면접자의 태도를 파악하는 데 한계가 있기 때문에, 지원자 음성을 주파수로 환산해 특성을 추출하고, Bidirectional LSTM을 활용해 훈련해 지원자 음성에 따른 6가지 감정을 추출했다. 셋째, 지원자의 발언 내용과 관련해 맥락적 의미를 파악해 지원자의 상태를 파악하기 위해, 음성을 STT(Speech-to-Text) 기법을 이용하여 텍스트로 변환하고, 사용 단어의 빈도를 분석하여 지원자의 언어 습관을 파악했다. 이와 함께, 지원자의 발언 내용에 대한 감정 분석을 위해 KoBERT 모델을 적용했으며, 지원자의 성격, 태도, 직무에 대한 이해도를 파악하기 위해 객관적인 평가지표를 제작하여 적용했다. 논문의 분석 결과 AI 면접의 다면적 평가시스템의 적절성과 관련해, 시각화 부분에서는 상당 부분 정확도가 객관적으로 입증되었다고 판단된다. 음성에서 감정분석 분야는 면접자가 제한된 시간에 모든 유형의 감정을 드러내지 않고, 또 유사한 톤의 말이 진행되다 보니 특정 감정을 나타내는 주파수가 다소 집중되는 현상이 나타났다. 마지막으로 자연어처리 영역은 면접자의 발언에서 나오는 말투, 특정 단어의 빈도수를 넘어, 전체적인 맥락과 느낌을 이해할 수 있는 자연어처리 분석모델의 필요성이 더욱 커졌음을 판단했다.

• Advanced Industrial SCIence
• /
• v.3 no.2
• /
• pp.8-16
• /
• 2024

This paper shows the system of drug classification, the goal of this is to foretell the apt drug for the patients based on their demographic and physiological traits. The dataset consists of various attributes like Age, Sex, BP (Blood Pressure), Cholesterol Level, and Na_to_K (Sodium to Potassium ratio), with the objective to determine the kind of drug being given. The models used in this paper are K-Nearest Neighbors (KNN), Logistic Regression and Random Forest. Further to fine-tune hyper parameters using 5-fold cross-validation, GridSearchCV was used and each model was trained and tested on the dataset. To assess the performance of each model both with and without hyper parameter tuning evaluation metrics like accuracy, confusion matrices, and classification reports were used and the accuracy of the models without GridSearchCV was 0.7, 0.875, 0.975 and with GridSearchCV was 0.75, 1.0, 0.975. According to GridSearchCV Logistic Regression is the most suitable model for drug classification among the three-model used followed by the K-Nearest Neighbors. Also, Na_to_K is an essential feature in predicting the outcome.

• Journal of The Korean Association For Science Education
• /
• v.27 no.9
• /
• pp.907-918
• /
• 2007

This study was conducted to explore key factors of expertise development of talented scientists who achieved outstanding research performance according to the stages of expertise development and dimensions of individual-domain-field. To fulfill the research purpose, 31 domestic scientists who were awarded major prizes in the field of science were interviewed in-depth from March to September, 2007. Stages of expertise development were analyzed in light of Csikszentmihalyi's IDFI (individual-domain-field interaction) model. Self-directed learning, multiple interests and finding strength, academic and liberal home environment, and meaningful encounter were major factors affecting expertise development in the exploration stage. In the beginner stage, independence, basic knowledge on major, and thirst for knowledge at university affected expertise development. Task commitment, finding flow, finding their field of interest and lifelong research topic, and mentor in formal education were the affecting factors in the competent stage. Finally, placing priority, communication skills, pioneering new domain, expansion of the domain, and evaluation and support system affected talented scientists' expertise development in the leading stage. The meaning of major patterns of expertise development were analyzed and described. Based on these analyses, educational implications for nurturing scientists were suggested.

• Journal of The Korean Association For Science Education
• /
• v.26 no.7
• /
• pp.775-789
• /
• 2006

The purpose of this study was to analyze the relevance of the school science contents that have been implemented for the 7-10th grade students in Korea since 2001. To fulfill the purpose of the study, we 1) analyzed the 7th national science curriculum of Korea, California science standards, the national science curriculum of England, Japanese national science curriculum, and current Korean and Japanese science textbooks, 2) conducted a nationwide survey in order to gather opinions from students, teachers, and science specialists. The main findings of this research were as follows: First, the number of topics presented at each grade level should be reduced and similar topics and themes should be integrated for students' deeper understanding. Second, science contents were excessive compared to the allotted teaching time and to foreign countries. Thus, the excessive overlapping and repetition of science contents should be avoided among the primary, middle and high school level, and the number of science concepts and activities should be reduced to an appropriate level considering time allotment for science classes, teachers' workload, laboratory conditions, etc. Third, to cope with the decreasing students' understanding and interests in science as school level and school year goes up, the science curriculum and textbooks should be developed to allow for student to learn science concepts by solving problems confronted in their daily lives. Fourth, a differentiated curriculum such as in-depth and supplementary course should be described not in science contents, but in teaching and learning strategy.

• Journal of the Korean Society of Radiology
• /
• v.18 no.2
• /
• pp.153-159
• /
• 2024

In X-ray projection, Unskilled radiologists become skilled through fail exam. This causes the patient to be exposed to unnecessary radiation. In this study, pre-position unskilled radiologic technologist presented ways to improve clinical proficiency. presented a skull lateral x-ray projection practice method using visual, spatial, and assistive devices. In addition, the accuracy and usefulness of the use of assistive devices were evaluated. When X-ray images were taken based on learning, the rotational spacing, which indicates image distortion, was 7.85 ± 1.45 mm and the tiliting spacing was 4.84 ± 0.5 mm. When practicing using visual aids, the rotational spacing is 4.4 ± 0.76 mm and the inclination spacing is 3.01 ± 0.87 mm. using a spatial compensation device, the rotational spacing is 5.2 ± 0.69 mm and the tiliting spacing is 3.33 ± 0.61 mm. Skull lateral X-ray Image distortion caused by empirical photography practice decreased by 5.4%, but image distortion caused by tilting increased by 1.2%. When practicing using a visual assistive devices, the degree of rotational spacing by 40.1% and the tiliting spacing decreased by 30.7% compared to the empirical x-ray exposure practice. When using spatial assistive devices, the rotation interval was reduced by 41.7% and the tilting interval by 23.7% compared to conventional empirical x-ray exposure practice. Therefore, if an unskilled radiologist practices using visual and spatial aids,the accuracy will be improved in skull lateral x-ray projection.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.19 no.1
• /
• pp.317-326
• /
• 2024

Biometric recognition is a technology that determines whether a person is identified by extracting information on a person's biometric and behavioral characteristics with a specific device. Cyber threats such as forgery, duplication, and hacking of biometric characteristics are increasing in the field of biometrics. In response, the security system is strengthened and complex, and it is becoming difficult for individuals to use. To this end, multiple biometric models are being studied. Existing studies have suggested feature fusion methods, but comparisons between feature fusion methods are insufficient. Therefore, in this paper, we compared and evaluated the fusion method of multiple biometric models using fingerprint, face, and iris images. VGG-16, ResNet-50, EfficientNet-B1, EfficientNet-B4, EfficientNet-B7, and Inception-v3 were used for feature extraction, and the fusion methods of 'Sensor-Level', 'Feature-Level', 'Score-Level', and 'Rank-Level' were compared and evaluated for feature fusion. As a result of the comparative evaluation, the EfficientNet-B7 model showed 98.51% accuracy and high stability in the 'Feature-Level' fusion method. However, because the EfficietnNet-B7 model is large in size, model lightweight studies are needed for biocharacteristic fusion.

• The Mathematical Education
• /
• v.63 no.1
• /
• pp.19-33
• /
• 2024

This study investigates longitudinal patterns in middle school students' mathematics interest and achievement using panel data from the 4th to 6th year of the Gyeonggi Education Panel Study. Results from the multivariate growth mixture model confirmed the existence of heterogeneous characteristics in the longitudinal trajectory of students' mathematics interest and achievement. Students were classified into four latent classes: a low-level class with weak interest and achievement, a high-level class with strong interest and achievement, a middlelevel-increasing class where interest and achievement rise with grade, and a middle-level-decreasing class where interest and achievement decline with grade. Each class exhibited distinct patterns in the change of interest and achievement. Moreover, an examination of the correlation between intercepts and slopes in the multivariate growth mixture model reveals a positive association between interest and achievement with respect to their initial values and growth rates. We further explore predictive variables influencing latent class assignment. The results indicated that students' educational ambition and time spent on private education positively affect mathematics interest and achievement, and the influence of prior learning varies based on its intensity. The perceived instruction method significantly impacts latent class assignment: teacher-centered instruction increases the likelihood of belonging to higher-level classes, while learner-centered instruction increases the likelihood of belonging to lower-level classes. This study has significant implications as it presents a new method for analyzing the longitudinal patterns of students' characteristics in mathematics education through the application of the multivariate growth mixture model.