• The Transactions of the Korea Information Processing Society
• /
• v.13 no.6
• /
• pp.284-290
• /
• 2024

Speech emotion recognition (SER) is a technique that is used to analyze the speaker's voice patterns, including vibration, intensity, and tone, to determine their emotional state. There has been an increase in interest in artificial intelligence (AI) techniques, which are now widely used in medicine, education, industry, and the military. Nevertheless, existing researchers have attained impressive results by utilizing acted-out speech from skilled actors in a controlled environment for various scenarios. In particular, there is a mismatch between acted and spontaneous speech since acted speech includes more explicit emotional expressions than spontaneous speech. For this reason, spontaneous speech-emotion recognition remains a challenging task. This paper aims to conduct emotion recognition and improve performance using spontaneous speech data. To this end, we implement deep learning-based speech emotion recognition using the VGG (Visual Geometry Group) after converting 1-dimensional audio signals into a 2-dimensional spectrogram image. The experimental evaluations are performed on the Korean spontaneous emotional speech database from AI-Hub, consisting of 7 emotions, i.e., joy, love, anger, fear, sadness, surprise, and neutral. As a result, we achieved an average accuracy of 83.5% and 73.0% for adults and young people using a time-frequency 2-dimension spectrogram, respectively. In conclusion, our findings demonstrated that the suggested framework outperformed current state-of-the-art techniques for spontaneous speech and showed a promising performance despite the difficulty in quantifying spontaneous speech emotional expression.

• Informatization Policy
• /
• v.31 no.2
• /
• pp.3-38
• /
• 2024

Hallucination is a significant barrier to the utilization of large-scale language models or multimodal models. In this study, we collected 654 computer science papers with "hallucination" in the abstract from arXiv from December 2022 to January 2024 following the advent of Chat GPT and conducted frequency analysis, knowledge network analysis, and literature review to explore the latest trends in hallucination research. The results showed that research in the fields of "Computation and Language," "Artificial Intelligence," "Computer Vision and Pattern Recognition," and "Machine Learning" were active. We then analyzed the research trends in the four major fields by focusing on the main authors and dividing them into data, hallucination detection, and hallucination mitigation. The main research trends included hallucination mitigation through supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF), inference enhancement via "chain of thought" (CoT), and growing interest in hallucination mitigation within the domain of multimodal AI. This study provides insights into the latest developments in hallucination research through a technology-oriented literature review. This study is expected to help subsequent research in both engineering and humanities and social sciences fields by understanding the latest trends in hallucination research.

• Journal of Oral Medicine and Pain
• /
• v.23 no.4
• /
• pp.419-432
• /
• 1998

The purpose of this Study was to examine the efficacy and compliance of a mandibular advancement device(MAD) according to the severity of sleep apnea in the snorers and obstructive sleep apnea patients. Fifty-four patients (45 males, 9 females, aged 20 - 68years ) who visited Seoul National Uiversity Dental Hospital(SNUDH) to seek for the treatment of snoring and sleep apnea were classified into four groups according to the results of the nocturnal polysomnography and they were instructed to wear MAD regularly which was designed to increase the size of the upper airway by advancing the mandible. The evaluation of the efficacy and compliance of the MAD according to the severity of apnea and the duration after the usage of MAD ( 1week, 1month, 3months, 6months, 12months) was made by using quesionnaires mad in Department of Oral Medicine and Oral diagnosis, SNUDH. The obtained results were as follows : 1. All subjects results were habitual snoreres and 43 patients(79.6%) complained the loudness of snoring that can be heard out of the room. 2. Apnea index(AI) of the total subjects was mean 29.4$\pm$26.9 and respiratory disturbance index(RDI)was mean 37.6$\pm$28.0. And there was nodifference in the efficacy and the compliances of MAD according to the severity of apnea. 3. The severityi of apnea by the questionnaires significantly corresponded with the results of nocturnal polysomnography, and this fact potentiated the diagnostic value of the questionnaire. 4. after the usage of MAD, there was significant improvement in the frequency of snoring, the loudness of snoring, frequency of apnea, daytime sleepiness nad the refreshment after sleep(p<0.001) regardless of the apnea index(AI) and respiratory distrubance index(RDI). 5. The degree of the satisfaction with MAD was mean 74.4$\pm$18.4% and that of the discomfort with the MAD was 31.4$\pm$19.6%. But there was no serious complication in occlusion and temporomandibular joint with the usage of MAD and the duration of the discomfort was mean 3.3$\pm$2.2 weeks. 6. Forty-one patients(75.9%) continued the usage of MAD but thirteen patients(24.1%) stopped the use of MAD because of the discomforts and insufficient results with it.

• Information Systems Review
• /
• v.22 no.1
• /
• pp.147-165
• /
• 2020

The sharing economy service is now spreading in various fields such as accommodation, cars and bicycles. In particular, bicycle-sharing service have become very popular around the world, and since September 2015, Seoul has been providing a bicycle-sharing service called 'Ttareungi'. However, the number of bicycles is unbalanced among rental stations continuously according to the user's bicycle use. In order to solve these problems, we employed social network analysis using Ttareungi data in Seoul, Korea. We analyzed degree centrality, closeness centrality, betweenness centrality and k-core. As a result, the degree centrality was found to be closely linked with bus or subway transfer center. Closeness centrality was found to be in an unbalanced departure and arrival frequency or poor public transport proximity. Betweenness centrality means where the frequency of departure and arrival occurs frequently. Finally, the k-core analysis showed that Mapo-gu was the most important group by time zone. Therefore, the results of this study may contribute to the planning of relocation and additional installation of bike rental station in Seoul.

• Journal of Navigation and Port Research
• /
• v.48 no.3
• /
• pp.137-145
• /
• 2024

Recently, there has been an increase in the use of Automatic Identification Systems. Class A AIS is used for ships engaged in international voyages, while Class B AIS is utilized for smaller vessels navigating domestic coastlines. AtoN AIS is used for aids to navigation, AIS is employed for search and rescue aircraft, and AIS-SART is widely used worldwide. Accordingly, in 2022, the Maritime Safety Committee(MSC) of the International Maritime Organization(IMO) revised the performance standards for the satellite emergency positioning radio beacon(EP IRB) to include AIS signals along with 121.5 MHz for aircraft, which has been used as a homing signal. It was recommended to use together as a homing signal, and from July 1, 2022, it was decided that AIS-EP IRB that satisfies the revised performance standards will replace the existing EP IRB. Consequently, starting from July 1, 2022, it was decided that AIS-EPIRB, which meets the revised performance standards, will replace the existing EP IRB. This paper aims to verify the feasibility of implementing AIS-EPIRB, which has not yet been developed domestically. To achieve this, a dedicated chipset for AIS was used to additionally implement frequency generation of 161.975 MHz and 162.025 MHz and GMSK modulation to satisfy the requirements.

• Information Systems Review
• /
• v.25 no.1
• /
• pp.75-91
• /
• 2023

Thanks to information technologies, sharing economy services offer a new way of consumption. Carsharing appeared as a novel type of service that transformed the conventional way of personal transportation, from owning a vehicle to using an on-demand service. Allowing users to use a vehicle without owning a car, carsharing provides various social benefits such as the reduction of resource allocation inefficiencies and the alleviation of transportation problems. To strengthen such positive aspects of carsharing service, it is essential to understand an individual's service usage pattern and reveal factors that affect users' reuse behavior. This study investigates the factors that have an influence on carsharing reuse of users applying RFMC (Recency, Frequency, Monetary, and Clumpiness) model, the popular model for understanding the reuse likelihood of customers. Using data from a leading carsharing service provider in South Korea, we empirically analyze the effect of RFMC on carsharing reuse behavior. The findings show that recency and monetary values are negatively related to reuse while frequency is positively related to carsharing service reuse. Moreover, the impact of recency and monetary value are more salient whereas the impact of frequency is smaller among users with higher clumpiness. Based on these findings, this study elaborates on theoretical and practical implications.

• Journal of the Korea Society for Simulation
• /
• v.31 no.3
• /
• pp.35-44
• /
• 2022

In this study, we analyze and forecast quantum computer technology trends. Previous research has been mainly focused on application fields centered on technology for quantum computer technology trends analysis. Therefore, this paper analyzes important quantum computer technologies and performs future signal detection and prediction, for a more market driven technical analysis and prediction. As analyzing words used in news articles to identify rapidly changing market changes and public interest. This paper extends conference presentation of Cha & Chang (2022). The research is conducted by collecting domestic news articles from 2019 to 2021. First, we organize the main keywords through text mining. Next, we explore future quantum computer technologies through analysis of Term Frequency - Inverse Document Frequency(TF-IDF), Key Issue Map(KIM), and Key Emergence Map (KEM). Finally, the relationship between future technologies and supply and demand is identified through random forests, decision trees, and correlation analysis. As results of the study, the interest in artificial intelligence was the highest in frequency analysis, keyword diffusion and visibility analysis. In terms of cyber-security, the rate of mention in news articles is getting overwhelmingly higher than that of other technologies. Quantum communication, resistant cryptography, and augmented reality also showed a high rate of increase in interest. These results show that the expectation is high for applying trend technology in the market. The results of this study can be applied to identifying areas of interest in the quantum computer market and establishing a response system related to technology investment.

• The Journal of the Korea Contents Association
• /
• v.20 no.3
• /
• pp.515-526
• /
• 2020

Digital healthcare is a combined area of medical field and IT and various information on digital healthcare is provided in social media. This study aims to find the research trend of digital healthcare by collecting and analyzing data related to digital healthcare through the social media. The data were collected from Naver and Daum's news and blogs from January 2008 to June 2019. Major keywords with high frequency were extracted and visualized with wordcloud and network analysis was used to analyze the relationship between major keywords. Research combining medical field and IT from 2008 to 2001, various convergence research based on medical field and IT from 2012 to 2015, convergence research that applied the 4th industrial revolution technologies such as big data, blockchain and AI were actively conducted from 2016 to June 2019.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2003.05a
• /
• pp.88-92
• /
• 2003

In this paper, the characteristics of the ZnO films deposited on AI bottom electrode and the temperature effects on the ZnO film growth are presented along with the fabrication and their evaluation of the film bulk acoustic wave resonator (FBAR) devices. All the films used in this work were deposited using a radio-frequency (RF) magnetron sputtering technique. Growth characteristics of the ZnO films are shown to have a strong dependence on the deposition temperatures ranged from room temperature to 35$0^{\circ}C$ regardless of the RF power applied for sputtering the ZnO target. In addition, according to the growth characteristics of the distinguishably different micro-crystal structures and the degree of the c-axis preferred orientation, the deposition temperatures can be divided into 3 temperature regions and 2 critical temperatures in-between. Overall, the ZnO films deposited at/below 20$0^{\circ}C$ are seen to have columnar grains with a highly preferred c-axis orientation where the full width at half maximum (FWHM) of X-ray diffraction rocking curve is 14$^{\circ}$. Based on the experimental findings, several FBAR devices were fabricated and measured. As a result, the FBAR devices show return loss of ~19.5dB at resonant frequency of ~2.05GHz.

• Journal of the Korean Society of Food Science and Nutrition
• /
• v.30 no.1
• /
• pp.168-174
• /
• 2001

This study was conducted to investigate correlation between soybean and their products consumption and risk factors for atherosclerosis in the healthyKorean adults. Health behaviors such as smoking, exercise, alcohol consumption and dietary patterns and nutrient intakes of 193 healthy adult subjects aged from 26 to 69 were assessed by using interview and semiquantitative food frequency questionnaire. The BMI, blood pressure and biochemical parameters of blood were examined as well as preferences for taste and family history of disease. Data were expressed as quartile according to soybean and their products consumption. The average daily soybean and their product consumption for 1st, 2nd, 3rd, and 4th percentile group were 36, 78, 112, and 182g, respectively. The more consumption of soybean and their products, the more intake of energy, protein, lipid, fiber, Ca, cholesterol as well as frequency of exercise, smoking and drinking. Serum TG, total cholesterol and LDL-cholesterol and AI as risk factors of atherosclerosis were positively correlated with smoking and drinking (p<0.05). Especially, serum TG was positively correlated with hypertension and BMI (p<0.01). But, no correlation between exercise, salty taste, meat preference, soybean products consumption and atherosclerosis risk factors was found, which means that life styles such as smoking and drinking rather than dietary habits might influence atherosclerosis in healthy adults. In conclusion, present soy products consumption should be increased by way of developing new generation soy products in order to exert anti-atherosclerotic effect by soybean in human.