• The Journal of the Convergence on Culture Technology
• /
• v.10 no.2
• /
• pp.517-524
• /
• 2024

This study aims to train and implement a deep learning model for the fusion of website creation and artificial intelligence, in the era known as the AI revolution following the launch of the ChatGPT service. The deep learning model was trained using 3,000 collected web page images, processed based on a system of component and layout classification. This process was divided into three stages. First, prior research on AI models was reviewed to select the most appropriate algorithm for the model we intended to implement. Second, suitable web page and paragraph images were collected, categorized, and processed. Third, the deep learning model was trained, and a serving interface was integrated to verify the actual outcomes of the model. This implemented model will be used to detect multiple paragraphs on a web page, analyzing the number of lines, elements, and features in each paragraph, and deriving meaningful data based on the classification system. This process is expected to evolve, enabling more precise analysis of web pages. Furthermore, it is anticipated that the development of precise analysis techniques will lay the groundwork for research into AI's capability to automatically generate perfect web pages.

• Annual Conference on Human and Language Technology
• /
• 2023.10a
• /
• pp.66-71
• /
• 2023

ChatGPT가 등장한 이후 다양한 거대 언어 모델(Large Language Model, LLM)이 등장하였고, 이러한 LLM을 목적에 맞게 파인튜닝하여 사용할 수 있게 되었다. 하지만 LLM을 새로 학습하는 것은 물론이고, 단순 튜닝만 하더라도 일반인은 시도하기 어려울 정도의 많은 컴퓨팅 자원이 필요하다. 본 연구에서는 공개된 LLM을 별도의 학습 없이 사용하여 zero-shot 프롬프팅으로 이진 분류 태스크에 대한 성능을 확인하고자 했다. 학습이나 추가적인 튜닝 없이도 기존 선학습 언어 모델들에 준하는 이진 분류 성능을 확인할 수 있었고, 성능이 좋은 LLM의 경우 분류 실패율이 낮고 일관적인 성능을 보여 상당히 높은 활용성을 확인하였다.

• Korean Journal of Food Science and Technology
• /
• v.29 no.1
• /
• pp.167-172
• /
• 1997

Effect of combined preparation (DWP715) containing Alaska pollack extract, maltol, ascorbic acid and nicotinamide on decreasing of blood alcohol was evaluated in human blood. Treatment of DWP715 prior to administration of 25% alcohol (100 mL) decreased alcohol concentration in blood and showed significant difference after 2 hours. The pharmacokinetic parameters such as area under the concentration-time curve (AUC), $C_{max},\;T_{max}\;and\;T_{1/2}$ were also decreased and delayed when compared with control values. Effects of DWP715 on anti-fatigue and anti-oxidation activities were also studied in the restraint stress model using various parameters (GOT, GPT, LDH values and organ weights) on mild condition and examined through the content of lipid peroxide induced by 2% $CCl_4$ in mouse livers. While GPT level, thymus and adrenal weight were not influenced by DWP715 dosing, LDH, GOT level and spleen weight used as a parameter against fatigue and stress states were recovered almost to the nomal level. Furthermore, lipid peroxidation due to $CCl_4$ was significantly inhibited by DWP715 treatment. These results suggest that DWP715 seems to metabolize the blood alcohol rapidly and to restore the damaged liver and fatigue conditions which was caused by alcohol metabolism to normal condition.

• Journal of the Korean Society for information Management
• /
• v.40 no.2
• /
• pp.115-135
• /
• 2023

The purpose of this study is to assess the effectiveness of using deep learning language models to extract references automatically and create a reference database for research reports in an efficient manner. Unlike academic journals, research reports present difficulties in automatically extracting references due to variations in formatting across institutions. In this study, we addressed this issue by introducing the task of separating references from non-reference phrases, in addition to the commonly used metadata extraction task for reference extraction. The study employed datasets that included various types of references, such as those from research reports of a particular institution, academic journals, and a combination of academic journal references and non-reference texts. Two deep learning language models, namely RoBERTa+CRF and ChatGPT, were compared to evaluate their performance in automatic extraction. They were used to extract metadata, categorize data types, and separate original text. The research findings showed that the deep learning language models were highly effective, achieving maximum F1-scores of 95.41% for metadata extraction and 98.91% for categorization of data types and separation of the original text. These results provide valuable insights into the use of deep learning language models and different types of datasets for constructing reference databases for research reports including both reference and non-reference texts.

• Annual Conference of KIPS
• /
• 2020.05a
• /
• pp.519-522
• /
• 2020

1989년에 WWW(World Wide Web)이 도입 되면서 세계적으로 인터넷의 보급이 시작되었다. 정보화 시대라고 알려진 3차 산업혁명 이후로 대량의 정보들이 소셜 미디어를 통하여 생산되었다. 소셜미디어는 2007년에 인터넷 사용자들 중 56%의 이용률을 보였지만 2008년 2분기에는 75%의 이용률로 증가함에 따라 대부분의 사용자들이 많이 사용하며 의존하게 되었다. 또한 소셜 미디어를 통해 발생 되는 데이터들을 이용하여 기업들은 이윤 창출을 할 수 있다. 하지만 이러한 소셜 미디어는 악의적인 목적을 통해 주가 조작, 정치적 선동 등을 할 수 있는 가짜 뉴스와 허위 정보들을 생성할 수 있으며 이에 따라 대책이 시급하다. 또한 가짜 뉴스는 사람이 글을 작성할 수도 있지만 최근 인공지능 기술의 발달에 따라 프로그램을 통해 자동적으로 생성 될 수도 있다. 본 논문에서는 이와 같은 실제 뉴스와 인공지능을 기반으로 한 뉴스를 분석한다. Kaggle에서 실제 뉴스 데이터를 수집하여 헤드라인을 OpenAI의 GPT-2 언어 모델을 통해 뉴럴 가짜 뉴스를 생성 하였다. 파이썬의 NLTK 모듈을 이용하여 전처리를 진행하였고 t-검정과 박스 플롯을 활용하여 분석을 진행하였다. 분석된 주요 속성들을 의사결정트리를 통해 모델 검증을 하였고 k-fold 교차검증을 통해 분류 모델을 평가하였다. 결과로 전체 분류 정확도 평균 89%의 성능을 보여주었다.

• Annual Conference on Human and Language Technology
• /
• 2022.10a
• /
• pp.531-534
• /
• 2022

신경망 기반 심층학습 기술은 대화처리 분야에서 대폭적인 성능 개선을 가져왔다. 특히 GPT-2와 같은 대규모 사전학습 언어모델을 백본 네트워크로 하고 특정 도메인 타스크 대화 데이터에 대해서 미세조정 방식으로 생성되는 종단형 대화모델의 경우, 해당 도메인 타스크에 대해서 높은 성능을 내고 있다. 하지만 이런 연구들은 대부분 하나의 도메인에 대해서만 초점을 맞출 뿐 싱글 모델로 두 개 이상의 도메인을 고려하고 있지는 않다. 특히 순차적인 미세 조정은 이전에 학습된 도메인에 대해서는 catastrophic forgetting 문제를 발생시킴으로써 해당 도메인 타스크에 대한 성능 하락이 불가피하다. 본 논문에서는 이러한 문제를 해결하기 위하여 MultiWoz 목적지향 대화 데이터에 오픈 도메인 칫챗 대화턴을 유사도에 기반하여 추가하는 데이터 증강 방식을 통해 사용자 입력 및 문맥에 따라 MultiWoz 목적지향 대화와 오픈 도메인 칫챗 대화를 함께 생성할 수 있도록 하였다. 또한 목적지향 대화와 오픈 도메인 칫챗 대화가 혼합된 대화에서의 시스템 응답 생성 성능을 평가하기 위하여 오픈 도메인 칫챗 대화턴을 수작업으로 추가한 확장된 MultiWoz 평가셋을 구축하였다.

• Annual Conference of KIPS
• /
• 2023.11a
• /
• pp.1033-1034
• /
• 2023

현대 사회에서 정신건강이 중요한 문제로 부상하고 있으나 국내 정신건강 서비스 이용률은 7.2%에 그친다. 코로나 발생 이후 이동성 제약 등의 요인에 따라 디지털 정신건강 관리 시장이 크게 성장할 것으로 보인다. 이에 본 논문에서는 AI 챗봇을 활용한 고민 상담을 통해 위로 및 제안을 제공하고, 대화 내용을 기반으로 영화를 추천하는 시스템을 제안한다. KoBert 모델을 이용하여 사용자의 감성을 분석하고, KoGPT 모델을 활용해 챗봇 응답을 생성한다.

• The Journal of Bigdata
• /
• v.7 no.2
• /
• pp.11-29
• /
• 2022

Pre-trained language models are the most important and widely used tools in natural language processing tasks. Since those have been pre-trained for a large amount of corpus, high performance can be expected even with fine-tuning learning using a small number of data. Since the elements necessary for implementation, such as a pre-trained tokenizer and a deep learning model including pre-trained weights, are distributed together, the cost and period of natural language processing has been greatly reduced. Transformer variants are the most representative pre-trained language models that provide these advantages. Those are being actively used in other fields such as computer vision and audio applications. In order to make it easier for researchers to understand the pre-trained language model and apply it to natural language processing tasks, this paper describes the definition of the language model and the pre-learning language model, and discusses the development process of the pre-trained language model and especially representative Transformer variants.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.24 no.4
• /
• pp.9-14
• /
• 2024

This study developed a voice recognition chatbot system to address depression and loneliness among the elderly in an aging society. The system utilizes the Whisper model, GPT 2.5, and XTTS2 to provide high-performance voice recognition, natural language processing, and text-to-speech conversion. Users can express their emotions and states and receive appropriate responses, with voice recognition functionality using familiar voices for comfort and reassurance. The UX/UI design considers the cognitive responses, visual impairments, and physical limitations of the smart senior generation, using high contrast colors and readable fonts for enhanced usability. This research is expected to improve the quality of life for the elderly through voice-based interfaces.

• The Korea Journal of Herbology
• /
• v.36 no.2
• /
• pp.31-42
• /
• 2021

Objectives : This study was undertaken to investigate the hepatoprotective effect of Spatholobi Caulis water extract (SC) to thioacetamide (TAA)-induced acute liver injury (ALI) in rats. Methods : The rats were injected intraperitoneally with TAA (200 mg/kg body weight) and orally administered SC (100 or 200 mg/kg b.w.) daily for 3 days. Liver biomarkers were assessed by serum glutamic oxaloacetic transaminase (GOT), glutamic pyruvic transaminase (GPT), and ammonia levels. Malondialdehyde (MDA) was measured both serum and liver tissue. In addition, nicotinamide adenine dinucleotide phosphate (NADPH) oxidase, anti-oxidant, and inflammation-related proteins were investigated by western blot analysis. Histological examination further confirmed though hematoxylin and eosin stain. Results : The SC treatment reduced liver function markers like GOT and GPT and also remarkably decreased ammonia level. Moreover, the elevated MDA level in TAA-induced group was significantly reduced by SC treatment. NADPH oxidase expression associated with oxidative stress including NOX2, NOX4, and p47phox markedly inhibited by SC administration. SC treatment exerted anti-oxidant effect through the increase of anti-oxidant enzyme including superoxide dismutase (SOD), Catalase, and heme oxygenase-1 (HO-1). The protein expressions of inflammatory cytokines such as tumor necrosis factor-�� (TNF-��), IL-6, and IL-1�� induced by nuclear factor-kappa B (NF-��B) activation were modulated through blocking the phosphorylation of inhibitor of nuclear factor ��B�� (I��B)��. SC treatment also improved histological alterations. Conclusion : These findings suggested that SC administration may be a potential candidate for the prevention or treatment of ALI.