• Journal of Korean Elementary Science Education
• /
• v.40 no.2
• /
• pp.162-174
• /
• 2021

The purpose of this study was to investigate the types of questions raised by pre-service elementary teachers when planning a science class utilizing smart technology and the characteristics of their lesson plans. For this purpose, lesson plans and questions written by the 96 pre-service teachers were collected. The results of this study can be summarized as follows: (1) Pre-service teachers used simulation apps, information offering apps, clicker evaluation apps, astronomical observation apps. Simulation apps and clicker evaluation apps were used the most in the introduction stage of the class, simulation apps in the development stage, and clicker evaluation apps in the finishing stage. (2) In the lesson plans, the activities that elementary school students use smart technology showed more than those used by teachers, and its characteristics were more prominent in the development stage of the class. (3) As for the content type of smart technology, experience type was the highest, followed by tool type and resource type. In comparison, there were relatively few interaction support types and learning opportunity extensions. (4) There were many cases in which pre-service teachers replaced elementary school students with virtual experiences using experience type instead of providing opportunities to experiment or experience directly. (5) Pre-service teachers asked various questions while planning science class utilizing smart technology, and a total of 25 question types appeared in 7 categories.

• KIPS Transactions on Software and Data Engineering
• /
• v.10 no.8
• /
• pp.319-330
• /
• 2021

Recently, in the field of open domain natural language question answering, multi-task, multi-hop question answering has been studied extensively. In this paper, we propose a novel deep neural network model using hierarchical graphs to answer effectively such multi-task, multi-hop questions. The proposed model extracts different levels of contextual information from multiple paragraphs using hierarchical graphs and graph neural networks, and then utilize them to predict answer type, supporting sentences and answer spans simultaneously. Conducting experiments with the HotpotQA benchmark dataset, we show high performance and positive effects of the proposed model.

• Journal of The Korean Association For Science Education
• /
• v.43 no.3
• /
• pp.253-263
• /
• 2023

This study explored the discrepancy between the text of a teacher's guide about straight and reflective light and the content knowledge of pre-service elementary teachers. A total of 455 questions and 543 answers generated by 279 pre-service elementary teachers after reading a 'Shadow and Mirror' unit in the teacher's guide were analyzed. The questions were classified according to the types of concepts and discrepancies, and the answers were analyzed for accuracy. The results of analyzing the concepts of questions revealed that the pre-service elementary teachers were most curious about the shadow in the straight concept, the mirror image in the reflection concept, and the light source in other concepts. The questions with a low correct answer rate due to incorrect- or non-answers, such as those concerning the superposition principle of light by reflection, the principle of experimental tools, and images by lenses, were only partially or not included in the teacher's guide. When the questions were classified according to the type of discrepancy, the frequency of questions due to knowledge deficit was higher than that due to knowledge clash. This demonstrates that the concepts that teachers need to know must be supplemented with the contents of the teacher's guide. Discrepancies due to knowledge clashes are often caused by conflicts between what is experienced in everyday life and what is presented in textbooks. Therefore, it is necessary to reduce the discrepancy between the texts of the teacher's guide and the knowledge of pre-service elementary teachers by including the differences between the actual context of everyday life and the context of the textbook in the teacher's guide.

• Journal of the Korean Academy of Child and Adolescent Psychiatry
• /
• v.15 no.1
• /
• pp.28-39
• /
• 2004

The author reviewed various psychological measures of attachment types and variation on developmental stage approaches. Also, the point of reliabilities and validities of the measures were reviewed. SSP and AQS introduced first in infancy. Cassidy-Marvin's attachment type system, Crittenden's PAA, Main & Cassidy's attachment system, Doll play techniques and Family drawing analyses of attachment types introduced secondly for preschooler. IPPA for parent and peer attachment introduced for Childhood/Adolescence. Finally, AAI and ASQ for romantic attachment explained for adulthood.

• Korean Journal of Childcare and Education
• /
• v.9 no.6
• /
• pp.243-265
• /
• 2013

The purpose of this study is to figure out how a child, as an active responder, responds to the characters of fantastic picture books by analyzing the child's questions and comments through the reading aloud approach. The subjects of this study are fifty-four children under five years old. Nine fantastic picture books are used as the study materials. The contents of the analysis are the frequency of children's questions and comments, the types of responses and the reasons of preferences to the characters. The results of the analysis are as follows: Firstly, the frequency of comments is three times higher than the frequency of questions. Secondly, the frequency of "evaluative questions" is the highest and "imaginative questions" is the next highest. The frequency of "transparent questions" and "personal questions" are comparatively low. Thirdly, most of the children answered that the reason of preference of the characters was "the character's appearance" and then "their subjective feeling to the character", "the character's role" and "the character's characteristics" in that order. Only one child answered that it was "the character's gender." This study will contribute to the planning and implementation of the strategies of reading picture books and to the strategic study to improve children's responses as well.

• Journal of the Korean Society for Library and Information Science
• /
• v.31 no.1
• /
• pp.153-184
• /
• 1997

The importance of WWW search services is increasing as Internet information resources explode. An evaluation of current 9 search services was first conducted by comparing descriptively the features concerning indexing, searching, and ranking of search results. Secondly, a couple of search queries were used to evaluate search performance of those services by the measures of retrieval effectiveness. the degree of overlap in searching sites, and the degree of similarity between services. In this experiment, Alta Vista, HotBot and Open Text Index showed better results for the retrieval effectiveness. The level of similarity among the 9 search services was extremely low.

• Journal of the Korean Society for information Management
• /
• v.31 no.4
• /
• pp.7-28
• /
• 2014

This study analyses the questions addressed to the collaborative digital reference service run by the National Library of Korea. The data consist of 661 question entries to the 'Ask a Librarian' service during first 6 months in 2014. Each entry includes average 1.17 questions, and 77.82% of the total questions are real reference in nature. Questions are analyzed by classification division, context of the questioner, desired end product, activities of librarians, and the resources used to respond them. Each category is subdivided and analyzed in detail. Results revealed interesting findings and problems, and suggestions for further endeavor are provided.

• Annual Conference on Human and Language Technology
• /
• 2019.10a
• /
• pp.213-218
• /
• 2019

기계독해는 자연어로 표현된 질문과 단락이 주어졌을 때, 해당 단락 내에 표현된 정답을 찾는 태스크이다. 최근 기계독해 태스크도 다른 자연어처리 태스크와 유사하게 BERT, XLNet, RoBERTa와 같이 사전에 학습한 언어모델을 이용하고 질문과 단락이 입력되었을 경우 정답의 경계를 추가 학습(fine-tuning)하는 방법이 우수한 성능을 보이고 있으며, 특히 KorQuAD v1.0 데이터셋에서 학습 및 평가하였을 경우 94% F1 이상의 높은 성능을 보이고 있다. 본 논문에서는 현재 최고 수준의 기계독해 기술이 학습셋과 유사한 평가셋이 아닌 일반적인 질문과 단락 쌍에 대해서 가지는 일반화 능력을 평가하고자 한다. 이를 위하여 첫번째로 한국어에 대해서 공개된 KorQuAD v1.0 데이터셋과 NIA v2017 데이터셋, 그리고 엑소브레인 과제에서 구축한 엑소브레인 v2018 데이터셋을 이용하여 데이터셋 간의 교차 평가를 수행하였다. 교차 평가결과, 각 데이터셋의 정답의 길이, 질문과 단락 사이의 오버랩 비율과 같은 데이터셋 통계와 일반화 성능이 서로 관련이 있음을 확인하였다. 다음으로 KorBERT 사전 학습 언어모델과 학습 가능한 기계독해 데이터 셋 21만 건 전체를 이용하여 학습한 기계독해 모델에 대해 블라인드 평가셋 평가를 수행하였다. 블라인드 평가로 일반분야에서 학습한 기계독해 모델의 법률분야 평가셋에서의 일반화 성능을 평가하고, 정답 단락을 읽고 질문을 생성하지 않고 질문을 먼저 생성한 후 정답 단락을 검색한 평가셋에서의 기계독해 성능을 평가하였다. 블라인드 평가 결과, 사전 학습 언어 모델을 사용하지 않은 기계독해 모델 대비 사전 학습 언어 모델을 사용하는 모델이 큰 폭의 일반화 성능을 보였으나, 정답의 길이가 길고 질문과 단락 사이 어휘 오버랩 비율이 낮은 평가셋에서는 아직 80%이하의 성능을 보임을 확인하였다. 본 논문의 실험 결과 기계 독해 태스크는 특성 상 질문과 정답 사이의 어휘 오버랩 및 정답의 길이에 따라 난이도 및 일반화 성능 차이가 발생함을 확인하였고, 일반적인 질문과 단락을 대상으로 하는 기계독해 모델 개발을 위해서는 다양한 유형의 평가셋에서 일반화 평가가 필요함을 확인하였다.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2015.05a
• /
• pp.260-260
• /
• 2015

에티오피아는 국제원조사업으로 SNNPR주 Gurage zone에 "안전한 식수 공급 사업"을 요청하였다. 이에 KOICA에서는 당해 지역내 국제원조사업 실시에 대한 성과를 평가하고자, ODA 사업 추진 전 지역주민의 "삶의 질"을 사전 평가하였다. 본 논고는 KOICA의 에티오피아 SNNPR주 Gurage zone "삶의 질" 평가 결과이다. "삶의 질" 분석을 위한 평가기준으로는 1. 현지 지역주민의 물이용 현황, 위생 및 환경 현황, 취수활동이 교육에 미치는 영향, 취수활동이 가정 경제 활동에 미치는 영향 등을 분석하고, 2. 식수공급사업 추진 시, 지역주민이 바라는 식수개발사업의 유형을 살펴보았다. 분석을 위한 자료 조사 방법으로는 설문 방법을 선택하였으며, 설문은 물 이용 현황, 질병 및 위생 관리 현황, 취수활동이 교육 및 가계 경제에 미치는 영향 그리고 지역주민이 바라는 식수사업의 방향 및 이유 등을 중심으로 구성하였다. 설문지 유형은 현황 관련 계량적 수치를 답하는 질문(계량적 평가)과 함께 지역주민이 현황에 관하여 인식하는 정도를 답하는 질문(비-계량적 평가, 리커트 5점 척도 이용)을 함께 구성하였다. 조사범위는 식수위생사업 범위에 있는 Gurage Zone 24개 마을, 301 가구, 400명의 주민을 대상으로 설문조사를 실시하였다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2009.11a
• /
• pp.945-946
• /
• 2009

현재의 컴퓨터 설문 시스템은 설문의 규칙이나 처리절차들을 하드코딩하는 방식으로 구현하기 때문에 설문이 변경되면 처리엔진을 수정해야 한다. 이것은 설문 처리엔진이 설문에 종속되어 있음을 의미한다. 본 논문에서는 이러한 약점을 해결하고자 능동문서 모델을 이용하여 설문으로부터 처리엔진을 독립시키는 방법을 고안하였다. 능동문서 기반의 설문지는 컨텐츠, 로직, 지식베이스(사용자 응답), 질의로 구성된다. 이 중에서 로직은 다양한 질문유형에 대한 처리 방법을 기술하는 것으로서 어떤 설문 설계자도 정의할 수 있어야 한다. 또한 로직은 직관적으로 서술할 수 있고 실행가능 하도록 XML 형식의 규칙 마크업 언어인 ERML로 표현하였다. ERML로 작성된 로직은 Prolog로 변환된 후 추론기에 등록되며, 사용자 응답에 따라 설문을 제어하고 처리한다. 마지막으로 몇 가지 질문유형을 ERML로 구현하고 설문 시스템(WINAD: The Web Interview System with Active Documents)에 적용한 실험 결과 설문으로부터 처리엔진을 독립시킬 수 있음을 확인했다.