• Title/Summary/Keyword: Bertopic

Search Result 27, Processing Time 0.021 seconds

Topic Modeling Insomnia Social Media Corpus using BERTopic and Building Automatic Deep Learning Classification Model (BERTopic을 활용한 불면증 소셜 데이터 토픽 모델링 및 불면증 경향 문헌 딥러닝 자동분류 모델 구축)

  • Ko, Young Soo;Lee, Soobin;Cha, Minjung;Kim, Seongdeok;Lee, Juhee;Han, Ji Yeong;Song, Min
    • Journal of the Korean Society for information Management
    • /
    • v.39 no.2
    • /
    • pp.111-129
    • /
    • 2022
  • Insomnia is a chronic disease in modern society, with the number of new patients increasing by more than 20% in the last 5 years. Insomnia is a serious disease that requires diagnosis and treatment because the individual and social problems that occur when there is a lack of sleep are serious and the triggers of insomnia are complex. This study collected 5,699 data from 'insomnia', a community on 'Reddit', a social media that freely expresses opinions. Based on the International Classification of Sleep Disorders ICSD-3 standard and the guidelines with the help of experts, the insomnia corpus was constructed by tagging them as insomnia tendency documents and non-insomnia tendency documents. Five deep learning language models (BERT, RoBERTa, ALBERT, ELECTRA, XLNet) were trained using the constructed insomnia corpus as training data. As a result of performance evaluation, RoBERTa showed the highest performance with an accuracy of 81.33%. In order to in-depth analysis of insomnia social data, topic modeling was performed using the newly emerged BERTopic method by supplementing the weaknesses of LDA, which is widely used in the past. As a result of the analysis, 8 subject groups ('Negative emotions', 'Advice and help and gratitude', 'Insomnia-related diseases', 'Sleeping pills', 'Exercise and eating habits', 'Physical characteristics', 'Activity characteristics', 'Environmental characteristics') could be confirmed. Users expressed negative emotions and sought help and advice from the Reddit insomnia community. In addition, they mentioned diseases related to insomnia, shared discourse on the use of sleeping pills, and expressed interest in exercise and eating habits. As insomnia-related characteristics, we found physical characteristics such as breathing, pregnancy, and heart, active characteristics such as zombies, hypnic jerk, and groggy, and environmental characteristics such as sunlight, blankets, temperature, and naps.

Research Trends Analysis on ESG Using Unsupervised Learning

  • Woo-Ryeong YANG;Hoe-Chang YANG
    • The Journal of Economics, Marketing and Management
    • /
    • v.11 no.3
    • /
    • pp.47-66
    • /
    • 2023
  • Purpose: The purpose of this study is to identify research trends related to ESG by domestic and overseas researchers so far, and to present research directions and clues for the possibility of applying ESG to Korean companies in the future and ESG practice through comparison of derived topics. Research design, data and methodology: In this study, as of October 20, 2022, after searching for the keyword 'ESG' in 'scienceON', 341 domestic papers with English abstracts and 1,173 overseas papers were extracted. For analysis, word frequency analysis, word co-occurrence frequency analysis, BERTopic, LDA, and OLS regression analysis were performed to confirm trends for each topic using Python 3.7. Results: As a result of word frequency analysis, It was found that words such as management, company, performance, and value were commonly used in both domestic and overseas papers. In domestic papers, words such as activity and responsibility, and in overseas papers, words such as sustainability, impact, and development were included in the top 20 words. As a result of analyzing the co-occurrence frequency of words, it was confirmed that domestic papers were related mainly to words such as company, management, and activity, and overseas papers were related to words such as investment, sustainability, and performance. As a result of topic modeling, 3 topics such as named ESG from the corporate perspective were derived for domestic papers, and a total of 7 topics such as named sustainable investment for overseas papers were derived. As a result of the annual trend analysis, each topic did not show a relatively increasing or decreasing tendency, confirming that all topics were neutral. Conclusions: The results of this study confirmed that although it is desirable that domestic papers have recently started research on consumers, the subject diversity is lower than that of overseas papers. Therefore, it is suggested that future research needs to approach various topics such as forecasting future risks related to ESG and corporate evaluation methods.

Analysis of Topics Related to Population Aging Using Natural Language Processing Techniques (자연어 처리 기술을 활용한 인구 고령화 관련 토픽 분석)

  • Hyunjung Park;Taemin Lee;Heuiseok Lim
    • Journal of Information Technology Services
    • /
    • v.23 no.1
    • /
    • pp.55-79
    • /
    • 2024
  • Korea, which is expected to enter a super-aged society in 2025, is facing the most worrisome crisis worldwide. Efforts are urgently required to examine problems and countermeasures from various angles and to improve the shortcomings. In this regard, from a new viewpoint, we intend to derive useful implications by applying the recent natural language processing techniques to online articles. More specifically, we derive three research questions: First, what topics are being reported in the online media and what is the public's response to them? Second, what is the relationship between these aging-related topics and individual happiness factors? Third, what are the strategic directions and implications for benchmarking discussed to solve the problem of population aging? To find answers to these, we collect Naver portal articles related to population aging and their classification categories, comments, and number of comments, including other numerical data. From the data, we firstly derive 33 topics with a semi-supervised BERTopic by reflecting article classification information that was not used in previous studies, conducting sentiment analysis of comments on them with a current open-source large language model. We also examine the relationship between the derived topics and personal happiness factors extended to Alderfer's ERG dimension, carrying out additional 3~4-gram keyword frequency analysis, trend analysis, text network analysis based on 3~4-gram keywords, etc. Through this multifaceted approach, we present diverse fresh insights from practical and theoretical perspectives.

Customer Voices in Telehealth: Constructing Positioning Maps from App Reviews (고객 리뷰를 통한 모바일 앱 서비스 포지셔닝 분석: 비대면 진료 앱을 중심으로)

  • Minjae Kim;Hong Joo Lee
    • Journal of Intelligence and Information Systems
    • /
    • v.29 no.4
    • /
    • pp.69-90
    • /
    • 2023
  • The purpose of this study is to evaluate the service attributes and consumer reactions of telemedicine apps in South Korea and visualize their differentiation by constructing positioning maps. We crawled 23,219 user reviews of 6 major telemedicine apps in Korea from the Google Play store. Topics were derived by BERTopic modeling, and sentiment scores for each topic were calculated through KoBERT sentiment analysis. As a result, five service characteristics in the application attribute category and three in the medical service category were derived. Based on this, a two-dimensional positioning map was constructed through principal component analysis. This study proposes an objective service evaluation method based on text mining, which has implications. In sum, this study combines empirical statistical methods and text mining techniques based on user review texts of telemedicine apps. It presents a system of service attribute elicitation, sentiment analysis, and product positioning. This can serve as an effective way to objectively diagnose the service quality and consumer responses of telemedicine applications.

A School-tailored High School Integrated Science Q&A Chatbot with Sentence-BERT: Development and One-Year Usage Analysis (인공지능 문장 분류 모델 Sentence-BERT 기반 학교 맞춤형 고등학교 통합과학 질문-답변 챗봇 -개발 및 1년간 사용 분석-)

  • Gyeongmo Min;Junehee Yoo
    • Journal of The Korean Association For Science Education
    • /
    • v.44 no.3
    • /
    • pp.231-248
    • /
    • 2024
  • This study developed a chatbot for first-year high school students, employing open-source software and the Korean Sentence-BERT model for AI-powered document classification. The chatbot utilizes the Sentence-BERT model to find the six most similar Q&A pairs to a student's query and presents them in a carousel format. The initial dataset, built from online resources, was refined and expanded based on student feedback and usability throughout over the operational period. By the end of the 2023 academic year, the chatbot integrated a total of 30,819 datasets and recorded 3,457 student interactions. Analysis revealed students' inclination to use the chatbot when prompted by teachers during classes and primarily during self-study sessions after school, with an average of 2.1 to 2.2 inquiries per session, mostly via mobile phones. Text mining identified student input terms encompassing not only science-related queries but also aspects of school life such as assessment scope. Topic modeling using BERTopic, based on Sentence-BERT, categorized 88% of student questions into 35 topics, shedding light on common student interests. A year-end survey confirmed the efficacy of the carousel format and the chatbot's role in addressing curiosities beyond integrated science learning objectives. This study underscores the importance of developing chatbots tailored for student use in public education and highlights their educational potential through long-term usage analysis.

Fine-tuning BERT-based NLP Models for Sentiment Analysis of Korean Reviews: Optimizing the sequence length (BERT 기반 자연어처리 모델의 미세 조정을 통한 한국어 리뷰 감성 분석: 입력 시퀀스 길이 최적화)

  • Sunga Hwang;Seyeon Park;Beakcheol Jang
    • Journal of Internet Computing and Services
    • /
    • v.25 no.4
    • /
    • pp.47-56
    • /
    • 2024
  • This paper proposes a method for fine-tuning BERT-based natural language processing models to perform sentiment analysis on Korean review data. By varying the input sequence length during this process and comparing the performance, we aim to explore the optimal performance according to the input sequence length. For this purpose, text review data collected from the clothing shopping platform M was utilized. Through web scraping, review data was collected. During the data preprocessing stage, positive and negative satisfaction scores were recalibrated to improve the accuracy of the analysis. Specifically, the GPT-4 API was used to reset the labels to reflect the actual sentiment of the review texts, and data imbalance issues were addressed by adjusting the data to 6:4 ratio. The reviews on the clothing shopping platform averaged about 12 tokens in length, and to provide the optimal model suitable for this, five BERT-based pre-trained models were used in the modeling stage, focusing on input sequence length and memory usage for performance comparison. The experimental results indicated that an input sequence length of 64 generally exhibited the most appropriate performance and memory usage. In particular, the KcELECTRA model showed optimal performance and memory usage at an input sequence length of 64, achieving higher than 92% accuracy and reliability in sentiment analysis of Korean review data. Furthermore, by utilizing BERTopic, we provide a Korean review sentiment analysis process that classifies new incoming review data by category and extracts sentiment scores for each category using the final constructed model.

Applying NIST AI Risk Management Framework: Case Study on NTIS Database Analysis Using MAP, MEASURE, MANAGE Approaches (NIST AI 위험 관리 프레임워크 적용: NTIS 데이터베이스 분석의 MAP, MEASURE, MANAGE 접근 사례 연구)

  • Jung Sun Lim;Seoung Hun, Bae;Taehoon Kwon
    • Journal of Korean Society of Industrial and Systems Engineering
    • /
    • v.47 no.2
    • /
    • pp.21-29
    • /
    • 2024
  • Fueled by international efforts towards AI standardization, including those by the European Commission, the United States, and international organizations, this study introduces a AI-driven framework for analyzing advancements in drone technology. Utilizing project data retrieved from the NTIS DB via the "drone" keyword, the framework employs a diverse toolkit of supervised learning methods (Keras MLP, XGboost, LightGBM, and CatBoost) enhanced by BERTopic (natural language analysis tool). This multifaceted approach ensures both comprehensive data quality evaluation and in-depth structural analysis of documents. Furthermore, a 6T-based classification method refines non-applicable data for year-on-year AI analysis, demonstrably improving accuracy as measured by accuracy metric. Utilizing AI's power, including GPT-4, this research unveils year-on-year trends in emerging keywords and employs them to generate detailed summaries, enabling efficient processing of large text datasets and offering an AI analysis system applicable to policy domains. Notably, this study not only advances methodologies aligned with AI Act standards but also lays the groundwork for responsible AI implementation through analysis of government research and development investments.