• Journal of Engineering Education Research
• /
• v.25 no.3
• /
• pp.35-44
• /
• 2022

Artificial intelligence is a core technology in the era of digital transformation, and as the technology level is advanced and used in various industries, its influence is growing in various fields, including social, ethical and legal issues. Therefore, it is time to raise social awareness on ethics of artificial intelligence as a prevention measure as well as improvement of laws and institutional systems related to artificial intelligence development. In this study, we analyzed unstructured data, typically text, such as online news articles and comments to confirm the degree of social awareness on ethics of artificial intelligence development. The analysis showed that the public intended to concentrate on specific issues such as "Human," "Robot," and "President" in 2018 to 2019, while the public has been interested in the use of personal information and gender conflics in 2020 to 2021.

• ETRI Journal
• /
• v.45 no.2
• /
• pp.318-328
• /
• 2023

Recently, embedded systems, such as mobile platforms, have multiple processing units that can operate in parallel, such as centralized processing units (CPUs) and neural processing units (NPUs). We can use deep-learning compilers to generate machine code optimized for these embedded systems from a deep neural network (DNN). However, the deep-learning compilers proposed so far generate codes that sequentially execute DNN operators on a single processing unit or parallel codes for graphic processing units (GPUs). In this study, we propose PartitionTuner, an operator scheduler for deep-learning compilers that supports multiple heterogeneous PUs including CPUs and NPUs. PartitionTuner can generate an operator-scheduling plan that uses all available PUs simultaneously to minimize overall DNN inference time. Operator scheduling is based on the analysis of DNN architecture and the performance profiles of individual and group operators measured on heterogeneous processing units. By the experiments for seven DNNs, PartitionTuner generates scheduling plans that perform 5.03% better than a static type-based operator-scheduling technique for SqueezeNet. In addition, PartitionTuner outperforms recent profiling-based operator-scheduling techniques for ResNet50, ResNet18, and SqueezeNet by 7.18%, 5.36%, and 2.73%, respectively.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2020.05a
• /
• pp.32-35
• /
• 2020

본 논문은 실시간 데이터 분석을 위한 컨테이너 가상화 기술 사용에 대한 효용성을 알아보기 위해 HDP 와 MapR 배포판에 포함된 Spark 를 도커라이징 전과 후 환경에 설치 후 HiBench 벤치마크 프로그램을 이용해 성능을 측정하였다. 그리고 성능 측정치에 대해 대응표본 t 검정을 이용하여 도커라이징 전과 후의 성능 차이가 있는지를 통계적으로 분석하였다. 분석 결과, HDP 는 도커라이징 전과 후에 대한 성능 차이가 있었지만 MapR 은 성능 차이가 없었다.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2021.10a
• /
• pp.530-532
• /
• 2021

This paper deals with AI education service that enables developers who have completed basic programming education to program in C/C++ language in order to learn big data and artificial intelligence. In addition, a customized development environment configuration system according to the development environment and how the user implements and tests artificial intelligence are explained. And also it has a function to check the effect on artificial intelligence through manipulation of various internal parameters. It is expected that it will be possible to develop artificial intelligence education services without language restrictions through networks in the future.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2022.07a
• /
• pp.3-5
• /
• 2022

혼합 중요도(mixed criticality) 시스템은 안전에 중요한 기능을 우선시하도록 하는 추가적인 안전 요구사항이 존재한다. 그러나 기존 실시간 시스템의 설계로는 이를 만족하지 못하며, 높은 중요도 태스크가 다른 낮은 중요도 태스크로부터 간섭을 받아 데드라인 미스와 같은 문제를 일으키는 중요도 역전(criticality inversion) 문제가 발생할 수 있다. 이러한 중요도 역전 문제를 해결하기 위해 주기 변환(period transformation) 기법을 사용할 수 있지만, 스케줄링 오버헤드의 증가로 인해 오히려 전반적인 태스크의 응답시간이 증가하는 또 다른 문제가 발생하게 된다. 본 논문에서는 주기 변환과 스케줄링 오버헤드 간의 트레이드오프 관계를 설명하고, 실시간 리눅스 시스템에서 해당 문제점을 재연한 후 주기 변환의 적정선을 분석하고자 실험을 진행하였다. 그 결과, 중요도 역전 문제를 해결하기 위한 주기 변환을 그대로 적용할 경우, 문맥 교환이 48.7% 증가 및 스케줄링 오버헤드가 28.7% 증가로 인해 전반적인 응답시간이 증가하여 데드라인 미스가 다수 발생하는 결과를 확인할 수 있었다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.9
• /
• pp.2334-2347
• /
• 2023

Metaverse services generate text data, data of ubiquitous computing, in real-time to analyze user emotions. Analysis of user emotions is an important task in metaverse services. This study aims to classify user sentiments using deep learning and pre-trained language models based on the transformer structure. Previous studies collected data from a single platform, whereas the current study incorporated the review data as "Metaverse" keyword from the YouTube and Google Play Store platforms for general utilization. As a result, the Bidirectional Encoder Representations from Transformers (BERT) and Robustly optimized BERT approach (RoBERTa) models using the soft voting mechanism achieved a highest accuracy of 88.57%. In addition, the area under the curve (AUC) score of the ensemble model comprising RoBERTa, BERT, and A Lite BERT (ALBERT) was 0.9458. The results demonstrate that the ensemble combined with the RoBERTa model exhibits good performance. Therefore, the RoBERTa model can be applied on platforms that provide metaverse services. The findings contribute to the advancement of natural language processing techniques in metaverse services, which are increasingly important in digital platforms and virtual environments. Overall, this study provides empirical evidence that sentiment analysis using deep learning and pre-trained language models is a promising approach to improving user experiences in metaverse services.

• The Journal of the Convergence on Culture Technology
• /
• v.9 no.6
• /
• pp.695-702
• /
• 2023

The purpose of this study is to develop an early childhood artificial intelligence education programs in kindergartens. In order to achieve the purpose of the study, a four-step program development procedure was taken: documentary analysis, program design and development, execution, and assessment. This study presented the purpose and goals, educational content, teaching-learning methods, and evaluation of the early childhood artificial intelligence education program in kindergarten. By implementing an artificial intelligence education in the process of developing the program, the practicality and utilization of the program were secured. This study is meaningful in that it derives practical support measures to apply and activate artificial intelligence education programs in the field of early childhood education.

• Journal of the Korean Society of Systems Engineering
• /
• v.19 no.2
• /
• pp.126-134
• /
• 2023

Drones are emerging as a new security threat, and the world is working to reduce them. Detection and identification are the most difficult and important parts of the anti-drone systems. Existing detection and identification methods each have their strengths and weaknesses, so complementary operations are required. Detection and identification performance in anti-drone systems can be improved through the use of artificial intelligence. This is because artificial intelligence can quickly analyze differences smaller than humans. There are three ways to utilize artificial intelligence. Through reinforcement learning-based physical control, noise and blur generated when the optical camera tracks the drone may be reduced, and tracking stability may be improved. The latest NeRF algorithm can be used to solve the problem of lack of enemy drone data. It is necessary to build a data network to utilize artificial intelligence. Through this, data can be efficiently collected and managed. In addition, model performance can be improved by regularly generating artificial intelligence learning data.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.11
• /
• pp.2966-2986
• /
• 2023

Climate change is a constant threat to human life, and it is important to understand the public perception of this issue. Previous studies examining climate change have been based on limited survey data. In this study, the authors used big data such as news articles and social media data, within which the authors selected specific keywords related to climate change. Using these natural language data, topic modeling was performed for discourse analysis regarding climate change based on various topics. In addition, before applying topic modeling, sentiment analysis was adjusted to discover the differences between discourses on climate change. Through this approach, discourses of positive and negative tendencies were classified. As a result, it was possible to identify the tendency of each document by extracting key words for the classified discourse. This study aims to prove that topic modeling is a useful methodology for exploring discourse on platforms with big data. Moreover, the reliability of the study was increased by performing topic modeling in consideration of objective indicators (i.e., coherence score, perplexity). Theoretically, based on the social amplification of risk framework (SARF), this study demonstrates that the diffusion of the agenda of climate change in public news media leads to personal anxiety and fear on social media.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.1
• /
• pp.49-57
• /
• 2022

Recently, various uses of artificial intelligence have been made possible through the deep artificial neural network structure of machine learning, demonstrating human-like capabilities. Unfortunately, the deep structure of the artificial neural network has not yet been accurately interpreted. This part is acting as anxiety and rejection of artificial intelligence. Among these problems, we solve the capability part of artificial neural networks. Calculate the size of the artificial neural network structure and calculate the size of data that the artificial neural network can process. The calculation method uses the group method used in mathematics to calculate the size of data and artificial neural networks using an order that can know the structure and size of the group. Through this, it is possible to know the capabilities of artificial neural networks, and to relieve anxiety about artificial intelligence. The size of the data and the deep artificial neural network are calculated and verified through numerical experiments.