• Journal of Medicine and Life Science
• /
• v.20 no.2
• /
• pp.53-59
• /
• 2023

This study explores the potential use of artificial intelligence (AI)-based services, specifically ChatGPT-3.5, in medical education. The application of this technology is acknowledged as a valuable tool for simulating authentic clinical scenarios and enhancing learners' diagnostic and communication skills. To construct a case, students received ChatGPT training using a clinical ethics casebook titled "Clinical Ethics Cases and Commentaries for Medical Students and Physicians." Subsequently, a role-play script was generated based on this training. The initial draft of the script was reviewed by two medical professors and was further optimized using ChatGPT-3.5. Consequently, a comprehensive role-play script, accurately reflecting real-world clinical situations, was successfully developed. This study demonstrates the potential for effectively integrating AI technology into medical education and provides a solution to overcome limitations in developing role-play scripts within conventional educational settings. However, the study acknowledges that AI cannot always generate flawless role-play scripts and recognizes the necessity of addressing these limitations and ethical concerns. The research explores both the potential and limitations of employing AI in the early stages of medical education, suggesting that future studies should focus on overcoming these limitations while further investigating the potential applications of AI in this field.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2023.07a
• /
• pp.541-544
• /
• 2023

LLM(Largescale Language Model)의 성능 향상을 위한 비용 효율적인 방법으로 ChatGPT, GPT-4와 같은 초거대 모델의 output에 대해 SLM(Small Language Model)을 finetune하는 방법이 주목받고 있다. 그러나, 이러한 접근법은 주로 범용적인 지시사항 모델을 위한 학습 방법으로 사용되며, 제한된 특정 도메인에서는 추가적인 성능 개선의 여지가 있다. 본 연구는 특정 도메인(Writing Assistant)에서의 성능 향상을 위한 새로운 방법인 Self-Guided Approach를 제안한다. Self-Guided Approach는 (1) LLM을 활용해 시드 데이터에 대해 도메인 특화된 metric(유용성, 관련성, 정확성, 세부사항의 수준별) 점수를 매기고, (2) 점수가 매겨진 데이터와 점수가 매겨지지 않은 데이터를 모두 활용하여 supervised 방식으로 SLM을 미세 조정한다. Vicuna에서 제안된 평가 방법인, GPT-4를 활용한 자동평가 프레임워크를 사용하여 Self-Guided Approach로 학습된 SLM의 성능을 평가하였다. 평가 결과 Self-Guided Approach가 Self-instruct, alpaca와 같이, 생성된 instruction 데이터에 튜닝하는 기존의 훈련 방법에 비해 성능이 향상됨을 확인했다. 다양한 스케일의 한국어 오픈 소스 LLM(Polyglot1.3B, PolyGlot3.8B, PolyGlot5.8B)에 대해서 Self-Guided Approach를 활용한 성능 개선을 확인했다. 평가는 GPT-4를 활용한 자동 평가를 진행했으며, Korean Novel Generation 도메인의 경우, 테스트 셋에서 4.547점에서 6.286점의 성능 향상이 발생했으며, Korean scenario Genration 도메인의 경우, 테스트 셋에서 4.038점에서 5.795 점의 성능 향상이 발생했으며, 다른 유사 도메인들에서도 비슷한 점수 향상을 확인했다. Self-Guided Approach의 활용을 통해 특정 도메인(Writing Assistant)에서의 SLM의 성능 개선 가능성을 확인했으며 이는 LLM에 비용부담을 크게 줄이면서도 제한된 도메인에서 성능을 유지하며, LLM을 활용한 응용 서비스에 있어 실질적인 도움을 제공할 수 있을 것으로 기대된다.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2023.07a
• /
• pp.3-4
• /
• 2023

자연어처리는 수많은 정보가 수집된 전자의무기록의 비정형 데이터에서 유의미한 정보나 패턴 등을 추출해 의료진의 의사결정을 지원하고, 환자에게 더 나은 진단이나 치료 등을 지원할 수 있어 큰 잠재력을 가지고 있다. 그러나 전자의무기록은 개인정보와 같은 민감한 정보가 다수 포함되어 있어 접근하기 어렵고, 이로 인해 충분한 양의 데이터를 확보하기 어렵다. 따라서 본 논문에서는 신뢰할 수 있는 의료 합성 텍스트를 생성하기 위해 ChatGPT 기반 의료 텍스트 증강 도구를 개발하였다. 이는 사용자가 입력한 실제 의료 텍스트로 의료 합성 데이터를 생성한다. 이를 위해, 적합한 프롬프트와 의료 텍스트에 대한 전처리 방법을 탐색하였다. ChatGPT 기반 의료 텍스트 증강 도구는 입력 텍스트의 핵심 키워드를 잘 유지하였고, 사실에 기반한 의료 합성 텍스트를 생성할 수 있다는 것을 확인할 수 있었다.

• Korean Journal of Soil Science and Fertilizer
• /
• v.7 no.4
• /
• pp.201-207
• /
• 1974

Some effects of nitrogen, phosphorus, potassium and respiratory inhibitor on growth of rice plant and activity of GOT, GPT and peroxidase for the rice root were investigated. Obtained results were summarized as follows: 1. Growth of rice root and plant applied with $NO_3$-N in culture solution was generally increased in the length and weight compared with that of $NH_4$-N plot. On the other hand, the GOT, GPT and peroxidase activity was more increased in the $NH_4$-N plot than in the $NO_3$-N plot. 2. Oxidative power of ${\alpha}$-naphthylamine in rice root was stronger in the $NO_3$-N plot than in the $NH_4$-N plot. 3. When rice plant was cultured in the medium which did not supplied nitrogen, phosphorus or potassium, respectively, GOT activity was more decreased than GPT activity, while peroxidase activity was increased mostly in the potassium-free plot. 4. When rice plant was cultivated in the culture solution added respiratory inhibitor, NaF, plant height was shortened in the order of nitrogen-free > $NH_4$-H > urea-N > $NO_3$-N plot, and GOT and GPT activity was also decreased in the order of nitrogen-free > $NH_4$-N > urea-N > $NO_3$-N plot.

• Annual Conference on Human and Language Technology
• /
• 2020.10a
• /
• pp.495-500
• /
• 2020

뉴스 기사는 반드시 객관적이고 넓은 시각으로 정보를 전달하지 않는다. 따라서 뉴스 기사를 기존의 추천 시스템과 같이 개인의 관심사나 사적 정보를 바탕으로 선별적으로 추천하는 것은 바람직하지 않다. 본 논문에서는 최대한 객관적으로 다양한 시각에서 비슷한 사건과 인물에 대해서 판단할 수 있도록 유사도 기반의 기사 추천 모델을 제시한다. 길이가 긴 문서 사이의 유사도를 측정하기 위해 GPT2 [1]언어 모델을 활용했다. 이 과정에서 단방향 디코더 모델인 GPT2 [1]의 단점을 추가 학습으로 개선했으며, 저장 공간의 효율과 핵심 문단 추출을 위해 BM25 [2]함수를 사용했다. 그리고 준 지도 학습 [3]을 통해 유사도 레이블링이 되어있지 않은 최신 뉴스 기사에 대해서도 자가 학습을 진행했으며, 이와 함께 길이가 긴 문단에 대해서도 효과적으로 학습할 수 있도록 문장 길이를 기준으로 3개의 단계로 나누어진 커리큘럼 학습 [4]방식을 적용했다.

• YAKHAK HOEJI
• /
• v.43 no.4
• /
• pp.516-525
• /
• 1999

Hedera rhombea (HR) has been used for treatments of hemorrage, chronic catarrh, jaundice, lithisis and convulsion. This study was done to isolate active compounds that have protective effect on liver damage. BuOH and EtOAc fractions of HR recovered serum glutamic pyruvic transaminase (GPT), glutamic oxaloacetic transaminase (GOT) and ${\gamma}-glutamyltranspeptidase$ (${\gamma}-GTP$) activities in CCl4 treated rats. We isolated 7 phenolic compounds from BuOH and EtOAc fractions, which were identified as 3-caffeoyl quinic acid, 3,4-di-O-caffeoyl quinic acid, 3,5-di-O-caffeoyl quinic acid, 4,5-di-O-caffeoyl quinic acid, caffeic acid, methyl 3,4-di-O-caffeoyl quinic acid and methyl 3,5-di-O-caffeoyl quinic acid by chemical and spectral analysis. These compounds reduced significantly serum GOT and GPT elevated by CCl4 treatment in rats, and 3-caffeoyl quinic acid, 3,5-di-O-caffeoyl quinic acid and caffeic acid also showed mild inhibitory activity against human immunodeficiency virus.

• Environmental Analysis Health and Toxicology
• /
• v.11 no.3_4
• /
• pp.11-16
• /
• 1996

Paraquat is a useful nonselective herbicide widely used throught the world. However accidental or intentional ingestion of the herbicide cause fatal pulmonary injuring. But there is not suitable antidote of paraquat intoxication and therapeutic agents now be used are not effective. So, in this study we intended to evaluate the inhibitory effects of acetylmannan from Dioscorea batalas on paraquat toxicity. 100mg/kg acetylmannan from wild or cultured Dioscorea bataras was administered orally to male SD rats for 3 days and the administration time interval was 24hours. After one hour of final administration, 50mg/kg paraquat was administered intraperitonially. After 24 hours, the biochemical parameters of blood and tissues were examined. In paraquat treated groups, sGPT, BUN, creatinine, ALP levels were increased by 2 to 4 times of normal values. However in acetylmannan from wild Dioscorea batatas treated groups, sGPT, BUN, creatinine, ALP levels in blood and lung tissue were significantly decreased to normal levels. In acetylmannan from cultured Dioscorea batatas treated groups, BUN, creatinine were significantly decreased to normal values, but not in sGPT, ALP levels. Therefore, we concluded that acetylmannan from wild Dioscorea batatas can be used as an. antidote of paraquat toxicity.

• International Journal of Advanced Culture Technology
• /
• v.11 no.3
• /
• pp.253-259
• /
• 2023

The increasing incorporation of AI in video storyboard creation has been observed recently. Traditionally, the production of storyboards requires significant time, cost, and specialized expertise. However, the integration of AI can amplify the efficiency of storyboard creation and enhance storytelling. In Korea, AiSAC stands at the forefront of AI-driven storyboard platforms, boasting the capability to generate realistic images built on open datasets foundations. Yet, a notable limitation is the difficulty in intricately conveying a director's vision within the storyboard. To address this challenge, we proposed the application of image generation features from ChatGPT and Midjourney to AiSAC. Through this research, we aimed to enhance the efficiency of storyboard production and refined the intricacy of expression, thereby facilitating advancements in the video production process.

• Journal of Intelligence and Information Systems
• /
• v.28 no.4
• /
• pp.347-364
• /
• 2022

Fraudulent companies or sellers strategically manipulate reviews to influence customers' purchase decisions; therefore, the reliability of reviews has become crucial for customer decision-making. Since customers increasingly rely on online reviews to search for more detailed information about products or services before purchasing, many researchers focus on detecting manipulated reviews. However, the main problem in detecting manipulated reviews is the difficulties with obtaining data with manipulated reviews to utilize machine learning techniques with sufficient data. Also, the number of manipulated reviews is insufficient compared with the number of non-manipulated reviews, so the class imbalance problem occurs. The class with fewer examples is under-represented and can hamper a model's accuracy, so machine learning methods suffer from the class imbalance problem and solving the class imbalance problem is important to build an accurate model for detecting manipulated reviews. Thus, we propose an OpenAI-based reviews generation model to solve the manipulated reviews imbalance problem, thereby enhancing the accuracy of manipulated reviews detection. In this research, we applied the novel autoregressive language model - GPT-3 to generate reviews based on manipulated reviews. Moreover, we found that applying GPT-3 model for oversampling manipulated reviews can recover a satisfactory portion of performance losses and shows better performance in classification (logit, decision tree, neural networks) than traditional oversampling models such as random oversampling and SMOTE.

• Journal of the East Asian Society of Dietary Life
• /
• v.4 no.2
• /
• pp.59-67
• /
• 1994

The present study was conducted to evaluate the hepatoprotective effect of puerariae Radix methanol extract on benzo(a) pyrene(B(a)P) - induced liver injuries in rats. In vitro experiment, primary cultured hepatocytes (5X105 cells/$m\ell$) were cultured for 20~24 hours after adding puerariae Radix mehtanol extract(32$\mu\textrm{g}$/$m\ell$) and B(a)P(50 uM). In vivo experiment, Puerariae Radix methanol extract(0.25 g/kg/day, per os) was administered for 7 days and B(a)P(0.1 mg/kg/day, intraperitoneally) was given after the last administration of extract. And then the hepatoprotective effect of Puerariae Radix methanol extract was investigated biochemically through in vitro and in vivo experiments. Namely, activities of enzymes (GOT, GPT and LDH) were measured and 3-(4,5-dimethylthiazol-2-yl)-2, 5-diphenyl tetrazolium bromide(MTT) assay were carried out in vitro cell culture study and GOT, GPT, LDH and ALP activities and HDL-cholesterol, total cholesterol and triglyceride contents were performed in vivo study. In vitro experiment, as a result of enzyme activity measurement(GOT, GPT and LDH) and MTT assay, GOT,GPT and LDH activities changed by B(a)P were recovered to normal levels and hepatocytes impaired by B(a)P were recovered to normal. In vivo experiment, Puerariae Radix methanol extract significantly decreased the enzyme activities(GOT, GPT, ALP and LDH in serum and GPT and ALP in tissue) and lipid contents in comparison to B(a)P-treated group.