• Title/Summary/Keyword: Document

Search Result 4,925, Processing Time 0.03 seconds

A Study of Pre-trained Language Models for Korean Language Generation (한국어 자연어생성에 적합한 사전훈련 언어모델 특성 연구)

  • Song, Minchae;Shin, Kyung-shik
    • Journal of Intelligence and Information Systems
    • /
    • v.28 no.4
    • /
    • pp.309-328
    • /
    • 2022
  • This study empirically analyzed a Korean pre-trained language models (PLMs) designed for natural language generation. The performance of two PLMs - BART and GPT - at the task of abstractive text summarization was compared. To investigate how performance depends on the characteristics of the inference data, ten different document types, containing six types of informational content and creation content, were considered. It was found that BART (which can both generate and understand natural language) performed better than GPT (which can only generate). Upon more detailed examination of the effect of inference data characteristics, the performance of GPT was found to be proportional to the length of the input text. However, even for the longest documents (with optimal GPT performance), BART still out-performed GPT, suggesting that the greatest influence on downstream performance is not the size of the training data or PLMs parameters but the structural suitability of the PLMs for the applied downstream task. The performance of different PLMs was also compared through analyzing parts of speech (POS) shares. BART's performance was inversely related to the proportion of prefixes, adjectives, adverbs and verbs but positively related to that of nouns. This result emphasizes the importance of taking the inference data's characteristics into account when fine-tuning a PLMs for its intended downstream task.

Analysis of Relationship between Housing Tenure and Birth in Newlywed Couples by Using Panel Data (패널자료를 이용한 신혼가구의 주택점유형태와 출산 관계 연구)

  • Shin, Hyungsub
    • Land and Housing Review
    • /
    • v.13 no.3
    • /
    • pp.39-55
    • /
    • 2022
  • In this study, we investigate the interrelationship between housing tenure and childbirth by exploiting the correlation probability effect method that accounts for household heterogeneity. Using the newlywed household panel from 2011 to 2022, we find that home ownership has a positive impact on childbirth in newlyweds. Specifically, newlywed households with housing tenure show a 6.2%p higher birth rate and a 5.7%p higher second childbirth than newlywed households living in rented houses. For the case of first childbirth, we employ the probability effect probit model since the endogeneity was not detected between housing tenure and birth rate. We document the differential effects of housing tenure on childbirth in that the first childbirth rate is higher for households without housing tenures. The negative effects on first childbirth could be attributed to the economic burden due to initial housing ownership, while housing tenure could eventually provide housing stability, leading to positive effects on more than one childbirth. Finally, we identify that households with childbirth over the last year show a 4.2%p and 3.9%p lower probabilities of housing tenure in the total sample and second childbirth sample, respectively. This suggests that the increased living cost due to childbirth could delay home ownership.

Consideration on supplementary matters when preparing radioactive waste self-disposal (방사성폐기물 자체처분 작성시 보완사항에 관한 고찰)

  • Lee, Kyung-Jae;Park, Sung-woo;Park, Young-Jae;Park, In-Sik
    • The Korean Journal of Nuclear Medicine Technology
    • /
    • v.26 no.1
    • /
    • pp.15-26
    • /
    • 2022
  • Purpose Recently, in the process of examining the self-disposal of radioactive waste by the Korea Institute of Nuclear Safety, it is difficult to reach the final approval process for self-disposal. In connection with this, we intend to increase the processing efficiency of self-disposal and strengthen safety by analyzing cases of recent supplementary matters. Materials and Methods From 2018 to 2021, we compare and review a supplementary requests that preparing the procedures and plans for the self-disposal of radioactive waste by 20 institutions. In this regard, based on the provisions of the Atomic Energy Safety Act, we derive a detailed proposals for the self-disposal of radioactive waste by arranging the review processing period calculation and supplementary requests that occurred during the review process. Results The representative supplementary requests of the Korea Institute of Nuclear Safety are the calculation of the storage period by type and nuclide of radioactive waste, the contents of the packaging container, the RASIS reporting method, the planned storage method for self-disposal, confirmation of the final disposal company, and the storage period of the waste filter Calculation, radioactive labeling, etc. And it is emphasized as important. Conclusion The expected effects of the guidelines reflecting the latest supplements include reduction of the time required for document preparation and increase of work processing efficiency, improvement of storage efficiency in the radioactive waste storage room, and economic cost reduction. If the radioactive waste self-disposal guideline presented in this study is applied to the field, it is thought that it will be helpful in improving the work efficiency of those who are experiencing difficulties.

Multi-source information integration framework using self-supervised learning-based language model (자기 지도 학습 기반의 언어 모델을 활용한 다출처 정보 통합 프레임워크)

  • Kim, Hanmin;Lee, Jeongbin;Park, Gyudong;Sohn, Mye
    • Journal of Internet Computing and Services
    • /
    • v.22 no.6
    • /
    • pp.141-150
    • /
    • 2021
  • Based on Artificial Intelligence technology, AI-enabled warfare is expected to become the main issue in the future warfare. Natural language processing technology is a core technology of AI technology, and it can significantly contribute to reducing the information burden of underrstanidng reports, information objects and intelligences written in natural language by commanders and staff. In this paper, we propose a Language model-based Multi-source Information Integration (LAMII) framework to reduce the information overload of commanders and support rapid decision-making. The proposed LAMII framework consists of the key steps of representation learning based on language models in self-supervsied way and document integration using autoencoders. In the first step, representation learning that can identify the similar relationship between two heterogeneous sentences is performed using the self-supervised learning technique. In the second step, using the learned model, documents that implies similar contents or topics from multiple sources are found and integrated. At this time, the autoencoder is used to measure the information redundancy of the sentences in order to remove the duplicate sentences. In order to prove the superiority of this paper, we conducted comparison experiments using the language models and the benchmark sets used to evaluate their performance. As a result of the experiment, it was demonstrated that the proposed LAMII framework can effectively predict the similar relationship between heterogeneous sentence compared to other language models.

A Case Study on the Application of AI-OCR for Data Transformation of Paper Records (종이기록 데이터화를 위한 AI-OCR 적용 사례연구)

  • Ahn, Sejin;Hwang, Hyunho;Yim, Jin Hee
    • Journal of the Korean Society for information Management
    • /
    • v.39 no.3
    • /
    • pp.165-193
    • /
    • 2022
  • It can be said that digital technology is at the center of the change in the modern work environment. In particular, in general public institutions that prove their work with records produced by business management systems and document production systems, the record management system is also the work environment itself. Gimpo City applied for the 2021 public cloud leading project of the National Information Society Agency (NIA) to proactively respond to the 4th industrial revolution technology era and implemented a public cloud-based AI-OCR technology enhancement project with 330 million won in support of 330 million won. Through this, it was converted into data beyond the limitations of non-electronic records limited to search and image viewing that depend on standardized index values. In addition, a 98% recognition rate was realized by applying a new technology called AI-OCR. Since digital technology has been used to improve work efficiency, productivity, development cost, and record management service levels of internal and external users, we would like to share the direction of enhancing expertise in the record management and implementation of work environment innovation.

Factors influencing success and safety of AED retrieval in out of hospital cardiac arrests in Singapore

  • NG, Jonathan Shen You;HO, Reuben Jia Shun;YU, Jae Yong;NG, Yih Yng
    • The Korean Journal of Emergency Medical Services
    • /
    • v.26 no.2
    • /
    • pp.97-111
    • /
    • 2022
  • Purpose: Automated External Defibrillator (AED) usage in out-of-hospital cardiac arrests (OHCAs) improves the survival of patients. In Singapore, public AEDs are protected by locked boxes with a 'break glass' mechanism to deter theft. Community responders have sustained injuries while breaking glass to retrieve AEDs. This unprecedented study aimed to elucidate the factors influencing successful retrieval of an AED and to document the prevalence of injuries. Methods: A survey was created and distributed. Participants were required to have responded to an OHCA in the past 12 months. Comparison tests were performed with the Fischer-Freeman-Halton Exact test or Pearson chi square test at 5% significance levels, and with multiple logistic regression with a logit link function. Results: Eighty-eight participants were eligible. The success of retrieving an AED was found not to be impacted by occupation, age, gender or time. Participants who responded to an OHCA because of activation by the myResponder App were more likely to retrieve an AED successfully. (AOR 11.111, 95% CI: 2.141-58.824) Conclusion: Use of the myResponder mobile application is associated with the greater success of retrieving an AED. Successful retrieval of an AED is not impacted by time, gender, age, or the occupation of the responder. Community responders in Singapore remain motivated to respond to Cardiac Arrests despite risk of injury.

Development of Scaffolding Strategies Model by Information Search Process (ISP) (정보탐색과정(ISP)에 의한 스캐폴딩 전략 모형 개발)

  • Jeong-Hoon Lim
    • Journal of Korean Library and Information Science Society
    • /
    • v.54 no.1
    • /
    • pp.143-165
    • /
    • 2023
  • This study aims to propose a scaffolding strategy that can be applied to the information search process by using Kuhlthau's ISP model, which presented a design and implementation strategy for the mediation role in the learning process. To this end, the relevant literature was reviewed to categorize scaffolding strategies, and impressions were collected from the students surveys after providing 150 middle school students in the Daejeon area with the project class to which the scaffolding strategy based on the ISP model was applied. The collected data were processed into a form suitable for analysis through data preprocessing for word frequencies to be extracted, and topic analysis was performed using STM (Structural Topic Modeling). First, after determining the optimal number of topics and extracting topics for each stage of the ISP model, the extracted topics were classified into three types: cognitive domain-macro perspective, cognitive domain-micro perspective, and emotional domain perspective. In this process, we focused on cognitive verbs and emotional verbs among words extracted through text mining, and presented a scaffolding strategy model related to each topic by reviewing representative document cases. Based on the results of this study, if an appropriate scaffolding strategy is provided at the ISP model stage, a positive effect on learners' self-directed task solving can be expected.

A Study on the Transformation and Issue of the Japanese-Chinese Word 'Library' (화제한어 '도서관' 명칭의 변용과 쟁점에 관한 연구)

  • Hee-Yoon Yoon
    • Journal of the Korean Society for Library and Information Science
    • /
    • v.57 no.1
    • /
    • pp.23-44
    • /
    • 2023
  • The word library(図書館) is a Japanese translation of the Western library or Bibliothek in the mid-Meiji period. This word has been accepted in Chinese(图书馆), Taiwan(圖書館), Korea(도서관), and Vietnam(Dđồ thư quán), which are Chinese-speaking countries. If so, when and who first introduced the term library to Japan and China? In Japan, the enlightenment thinker Fukuzawa's 『Seiyo Jijo, 1866』 is regarded as the first document to introduce the Western library, and in China, the article published in 『Qing Yi Bao, 1896』 by the reformed thinker Liang Qichao referred to as the first example. Therefore, this study traced and demonstrated the time and person in which the word library appeared, focusing on modern dictionaries, books, translations, papers, and newspaper articles that were introduced in both countries. As a result, the theory of the introduction to Fukuzawa in 1866 is wrong because Western libraries are described in various terms in many diaries and dictionaries, including Motoki's 『An English Japanese Dictionary of the Spoken Language, 1814』. Also, in China, the theory of introduction of Liang Qichao in 1896 is not true because the term library first appeared in Ryu Jeong-dam's 『A Dictionary of Loan Words and Hybrid Words in Chinese, 1884』. In the same context, it is necessary to trace and argue the history of the first use of the term library in Korea and the name of the first library in Korea established by the Busan Branch of the Japan Hongdo Association in 1901.

A Study on the Blockchain based Frequency Allocation Process for Private 5G (블록체인 기반 5G 특화망 주파수 할당 프로세스 연구)

  • Won-Seok Yoo;Won-Cheol Lee
    • The Journal of Korea Institute of Information, Electronics, and Communication Technology
    • /
    • v.16 no.1
    • /
    • pp.24-32
    • /
    • 2023
  • The current Private 5G use procedure goes through the step of application examination, use and usage inspection, and can be divided in to application, examination step as a procedure before frequency allocation, and use, usage inspection step as a procedure after frequency allocation. Various types of documents are required to apply for a Private 5G, and due to the document screening process and radio station inspection for using Private 5G frequencies, the procedure for Private 5G applicants to use Private 5G is complicated and takes a considerable amount of time. In this paper, we proposed Frequency Allocation Process for Private 5G using a blockchain platform, which is fast and simplified than the current procedure. Through the use of a blockchain platform and NFT (Non-Fungible Token), reliability and integrity of the data required in the frequency allocation process were secured, and security of frequency usage information was maintained and a reliable Private 5G frequency allocation process was established. Also by applying the RPA system that minimizes human intervention, fairness was secured in the process of allocating Private 5G. Finally, the frequency allocation process of Private 5G based on the Ethereum blockchain was performed though a simulation.

A Study on the Improvement of Safety and Health Activities in the Construction Contractor (Public Institutions) (건설공사 발주처(공공기관) 안전보건활동 수준향상에 관한 연구)

  • Ji-Hwan Moon
    • Journal of the Society of Disaster Information
    • /
    • v.19 no.3
    • /
    • pp.624-633
    • /
    • 2023
  • Purpose: It Vas intended to identify problems and derive improvement plans by grasping the current status of safety management of public institutions among construction Vork orders. Method: By comparing the disaster status of public institutions compared to the total construction Vork, the analysis Vas conducted based on the results of the evaluation of the level of safety activities of public institutions Vith a high disaster rate and the results of actual consulting. Result: As a result of comparing and analyzing the current status of safety management of public institutions, the current status and problems of safety management in public institutions Vith a high accident rate Vere similarly discovered. Safety management organizations, document management systems, safety management systems, and risk assessment activities are operated Vithout reflecting the size and characteristics of the organization, so improvement in the relevant field is needed. Conclusion: Safety-related professionals and organizations should be formed according to the size of construction orders, and responsibility and authority should be clearly assigned. Since risk assessment is conducted formally to prepare a safety and health ledger, it is necessary to derive risk factors to prevent safety accidents for the actual construction. It is expected that the level of safety activities can be improved if it is improved by reflecting the size and characteristics of public institutions.