• The Journal of the Korea Contents Association
• /
• v.20 no.10
• /
• pp.89-99
• /
• 2020

In this study, we tried to produce moving-image works through sentiment analysis of music. First, Google natural language API was used for the sentiment analysis of lyrics, then the result was applied to the image visualization rules. In prior engineering researches, text-based sentiment analysis has been conducted to understand users' emotions and attitudes by analyzing users' comments and reviews in social media. In this study, the data was used as a material for the creation of artworks so that it could be used for aesthetic expressions. From the machine's point of view, emotions are substituted with numbers, so there is a limit to normalization and standardization. Therefore, we tried to overcome these limitations by linking the results of sentiment analysis of lyrics data with the rules of formative elements in visual arts. This study aims to transform existing traditional art works such as literature, music, painting, and dance to a new form of arts based on the viewpoint of the machine, while reflecting the current era in which artificial intelligence even attempts to create artworks that are advanced mental products of human beings. In addition, it is expected that it will be expanded to an educational platform that facilitates creative activities, psychological analysis, and communication for people with developmental disabilities who have difficulty expressing emotions.

• Science of Emotion and Sensibility
• /
• v.18 no.2
• /
• pp.75-84
• /
• 2015

Although various methods have been proposed and utilized for video reality improvement, use of olfaction still remains at a rudimentary stage. Previous research reported reality improvement effect of some scents when a video displayed specific objects whose odors matched to the scents provided. In addition, another study showed that provision of scents that correspond to the prevailing color of a video improves sense of immersion. However, the above studies have clear limitations because not all videos have specific odor or obvious color. Assuming, in this study, that sensibility-based scent provision in the absence of main odor or color will increase sense of reality, the present study aimed at building adjective profiles of various scents that convey different sensibilities. To this end, in Experiments 1 and 2, we collected a set of adjectives appropriate for description of scents, and in Experiment 3, we built profiles of 16 scents using 30 adjectives. In addition, we grouped scents of similar sensibilities using cluster analysis. These results could be used not only for video reality improvement but also for the purposes of emphasizing product concepts or store positioning, etc.

• Journal of Internet Computing and Services
• /
• v.24 no.4
• /
• pp.127-135
• /
• 2023

This research paper explores the application of FinBERT, a variational BERT-based model pre-trained on financial domain, for sentiment analysis in the financial domain while focusing on the process of identifying suitable training data and hyperparameters. Our goal is to offer a comprehensive guide on effectively utilizing the FinBERT model for accurate sentiment analysis by employing various datasets and fine-tuning hyperparameters. We outline the architecture and workflow of the proposed approach for fine-tuning the FinBERT model in this study, emphasizing the performance of various datasets and hyperparameters for sentiment analysis tasks. Additionally, we verify the reliability of GPT-3 as a suitable annotator by using it for sentiment labeling tasks. Our results show that the fine-tuned FinBERT model excels across a range of datasets and that the optimal combination is a learning rate of 5e-5 and a batch size of 64, which perform consistently well across all datasets. Furthermore, based on the significant performance improvement of the FinBERT model with our Twitter data in general domain compared to our news data in general domain, we also express uncertainty about the model being further pre-trained only on financial news data. We simplify the complex process of determining the optimal approach to the FinBERT model and provide guidelines for selecting additional training datasets and hyperparameters within the fine-tuning process of financial sentiment analysis models.

• Information and Communications Magazine
• /
• v.26 no.1
• /
• pp.25-30
• /
• 2009

IT산업은 '97년 외환위기를 기점으로 고성장을 거듭하였고, 세계 10위권 규모의 경제 강국으로 성장하는데 있어 국내 산업의 대표적인 성장동력으로서 자리 매김하였다. 하지만, 최근 성장률이 예전만 못하다는 우려의 목소리와 함께 IT가 한 단계 더 업그레이드하기 위해서는 자신의 허물을 과감히 벗는 용기가 필요하다는 것이 일반적인 시각이다. 이러한 흐름속에서 기존 IT전략과 차별화된 New IT 발전방안을 수립하게 되었다. 지금까지의 IT전략이 IT 고도화 추진을 위한 IT 중심 발전 전략이었다면, New IT전략은 IT기반 융합산업, IT융합 신산업, Next IT산업 등3대 성장축간 시너지 창출이 가능한 IT로의 수렴과 확산이 핵심이라고 할 수 있다. IT기반 융합산업의 최종 목표는 전통산업과 IT가 만나 자동차, 조선, 의료, 국방, 건설, 섬유, 기계항공 등 7개 주력기간산업의 고부가가치화 및 초일류화를 우선 실현하는 것이다. IT융합 신산업의 최종 목표는 나노(NT), 바이오(BT), 인지기술(CT) 등 비(非)IT와 교감을 통하여 IT가 에너지, 환경, 건강 등 사회적 문제를 해결하고, 녹색성장 추진의 핵심인 5대 신산업을 창출하는데 있다. 5대 신산업으로는 Green IT산업, Welfare-Infra산업, 감성조명산업, 인지단말산업, THz응용산업 등을 꼽을 수 있다. Next IT산업은 기존 14대 IT 분야를 'ETRI 비전 2020' 등 미래 청사진을 바탕으로 장기적 관점에서 재설계한 것이며, 최종 목표는 TDX, 4M D램, CDMA, 와이브로, DMB, NoLA 등 IT 강국 계보를 이어갈 4G, 미래인터넷, Smart Radio, 실감미디어, 웹3.0, 투명전자소자 등 미래 유망 핵심원천기술을 확보하고, IT 경쟁력을 강화하는 것이다.

• Science of Emotion and Sensibility
• /
• v.26 no.4
• /
• pp.29-40
• /
• 2023

Anxiety is known to upset the balance of the attentional system and prioritize the stimulus-driven system over the goal-directed system; however, self-instruction induces goal-directed behavior with the self-regulation effect. This study verified the effects of meteorological and self-instruction conditions on pilot anxiety and flight task performance for in-service pilots in a virtual reality environment. The meteorological conditions were divided into visual meteorological and very low visibility conditions, and the flight tasks were conducted by varying whether or not self-instruction was performed. The experiment results reveal that anxiety and heart rate were higher, and the performance of the flight task was lower in the very low visibility condition. However, anxiety and heart rate were lower, and the performance of the flight task was higher in the self-instruction condition. This result suggests that accidents due to difficulty in flight may increase because of anxiety, but such accidents may decrease because of flight performance improvement by self-instruction.

• Science of Emotion and Sensibility
• /
• v.19 no.2
• /
• pp.55-66
• /
• 2016

This study was aimed to characterize the color of cotton fabric dyed with three different natural resources such as persimmon, citrus, and brown algae and to propose new trend color themes for fashion color planning for newborn and toddler wear. A variety of color shades by dyeing cotton fabric with persimmon, citrus, and brown algae respectively was obtained and each color was also determined if it could be matched with one of Pantone TPX considering authorized color trends for 2016 Spring/Summer. Finally a group of new trend color theme for 2016S/S newborn and toddler wear were proposed for naturally dyed cotton fabric. As results, the three natural dyeing resources gave color gamut with hue ranging from Purple Blue to Green Yellow and with tones including pale, light grayish, soft, grayish, and dull by their single and compound dyeing. A total of 23 colors matched to Pantone TPX were considered for the fashion color planning by natural dyeing. The selected natural colors were grouped as three different color ways and thereafter the color ways were differentiated in terms of representative color sensibility by using subjective evaluation. Finally three characterized color themes were proposed as 'Serenity', 'Juicy', and 'Fancy', each of which has tried to express differentiated feeling of each natural resources for dyeing, to follow up to global color trends, and to contribute to newborn and toddler wear's own requirements and marketability. These results suggest that natural dyeing colors could be applied into fashion color planning in current fashion industries in order to produce more sensible and emotional design of fashion goods using natural dyeing.

• Science of Emotion and Sensibility
• /
• v.20 no.3
• /
• pp.109-118
• /
• 2017

This study examined the effects of button response to probe and irrelevant stimuli on P300 amplitude and lie detection rate in P300-based concealed information test. Participants underwent the P300-based concealed information test (P300 CIT) in two conditions. In one button condition participants were instructed to press the left mouse button only when the target was present. In two button condition, they were asked to press the left mouse button for target and right button for non-target. The results showed that the response time to target stimulus was not significantly different between the two conditions, and the response time to the probe stimulus was significantly longer than the irrelevant stimulus. The P300 amplitudes for the probe and irrelevant stimulus were all smaller in one button condition compared to two button condition. However, the P300 amplitude difference between the probe stimulus and the irrelevant stimulus did not show a significant difference in the two experimental conditions, and the lie detection rate did not differ significantly between the two conditions. Based on these findings, the effect of button response on P300 CIT with a short inter-stimulus interval was discussed.

• Science of Emotion and Sensibility
• /
• v.21 no.2
• /
• pp.73-88
• /
• 2018

The present study examined how processing features of visual information near the hand would affect task switching. Recent studies reported enhanced cognitive control of visual information presented the near hands. To investigate the enhancement of cognitive control based on the relationship between hand proximity and attention, we implemented 2 experiments. In the task switching performance experiment, the hand proximity effect depended on modality of cue and target. The first experiment showed that stimuli near the hand received greater cognitive control than stimuli far from the hand, resulting in smaller switch cost. The result could rule out the feature-binding problem, which identifies reduced switch cost as the cause instead of hand proximity. Our results show that hand proximity actually reduced switch cost. In the second experiment, we examined the effects of hand nearness, modality, and their interaction on switch cost. In task switching, the target was always visual, and the cue was presented either visually or auditorily. In addition, we manipulated the cue-target interval to observe the preparation effect of cue. The results showed that a visual cue near the hand reduced switch cost by shortening task preparation time. Also, modality switching between an auditory cue and visual target was remarkable in a hand-near condition. The results for the visual cue could be interpreted as a benefit of rapid visual attention orienting. On the other hand, the results for the auditory cue could be interpreted as the cost of interference of modality switching by slower attentional disengagement of stimuli near the hands. Finally, modulation of switch cost by attention induced by hand nearness was discussed.

• Journal of Internet Computing and Services
• /
• v.25 no.1
• /
• pp.57-67
• /
• 2024

This research investigates training methods for large language models to accurately identify sentiments and comprehend information about increasing and decreasing fluctuations in the financial domain. The main goal is to identify suitable datasets that enable these models to effectively understand expressions related to financial increases and decreases. For this purpose, we selected sentences from Wall Street Journal that included relevant financial terms and sentences generated by GPT-3.5-turbo-1106 for post-training. We assessed the impact of these datasets on language model performance using Financial PhraseBank, a benchmark dataset for financial sentiment analysis. Our findings demonstrate that post-training FinBERT, a model specialized in finance, outperformed the similarly post-trained BERT, a general domain model. Moreover, post-training with actual financial news proved to be more effective than using generated sentences, though in scenarios requiring higher generalization, models trained on generated sentences performed better. This suggests that aligning the model's domain with the domain of the area intended for improvement and choosing the right dataset are crucial for enhancing a language model's understanding and sentiment prediction accuracy. These results offer a methodology for optimizing language model performance in financial sentiment analysis tasks and suggest future research directions for more nuanced language understanding and sentiment analysis in finance. This research provides valuable insights not only for the financial sector but also for language model training across various domains.

• Journal of Service Research and Studies
• /
• v.14 no.1
• /
• pp.77-90
• /
• 2024

This research aims to investigate methods of leveraging generative artificial intelligence in service sectors where consumer sentiment and experience are paramount, focusing on minimizing hallucination phenomena during usage and developing strategic services tailored to consumer sentiment and experiences. To this end, the study examined both mechanical approaches and user-generated prompts, experimenting with factors such as business item definition, provision of persona characteristics, examples and context-specific imperative verbs, and the specification of output formats and tone concepts. The research explores how generative AI can contribute to enhancing the accuracy of personalized content and user satisfaction. Moreover, these approaches play a crucial role in addressing issues related to hallucination phenomena that may arise when applying generative AI in real services, contributing to consumer service innovation through generative AI. The findings demonstrate the significant role generative AI can play in richly interpreting consumer sentiment and experiences, broadening the potential for application across various industry sectors and suggesting new directions for consumer sentiment and experience strategies beyond technological advancements. However, as this research is based on the relatively novel field of generative AI technology, there are many areas where it falls short. Future studies need to explore the generalizability of research factors and the conditional effects in more diverse industrial settings. Additionally, with the rapid advancement of AI technology, continuous research into new forms of hallucination symptoms and the development of new strategies to address them will be necessary.