• Annual Conference on Human and Language Technology
• /
• 2014.10a
• /
• pp.112-116
• /
• 2014

서답형 문항은 학생들의 종합적인 사고능력을 판단하는데 매우 유용하지만 채점할 때, 시간과 비용이 매우 많이 소요되고 채점자의 공정성을 확보해야 하는 어려움이 있다. 이러한 문제를 개선하기 위해 본 논문에서는 서답형 문항에 대한 자동채점 시스템을 제안한다. 본 논문에서 제안하는 시스템은 크게 언어 처리 단계와 채점 단계로 나뉜다. 첫 번째로 언어 처리 단계에서는 형태소 분석과 같은 한국어 정보처리 시스템을 이용하여 학생들의 답안을 분석한다. 두 번째로 채점 단계를 진행하는데 이 단계는 아래와 같은 순서로 진행된다. 1) 첫 번째 단계에서 분석 결과가 완전히 일치하는 답안들을 하나의 유형으로 간주하여 각 유형에 속한 답안의 빈도수가 높은 순서대로 정렬하여 인간 채점자가 고빈도 학생 답안을 수동으로 채점한다. 2) 현재까지 채점된 결과와 모범답안을 학습말뭉치로 간주하여 자질 추출 및 자질 가중치 학습을 수행한다. 3) 2)의 학습 결과를 토대로 미채점 답안들을 군집화하여 분류한다. 4) 분류된 결과 중에서 신뢰성이 높은 채점 답안에 대해서 인간 채점자가 확인하고 학습말뭉치에 추가한다. 5) 이와 같은 방법으로 미채점 답안이 존재하지 않을 때까지 반복한다. 제안된 시스템을 평가하기 위해서 2013년 학업성취도 평가의 사회(중3) 및 국어(고2) 과목의 서답형 문항을 사용하였다. 각 과목에서 1000개의 학생 답안을 추출하여 채점시간과 정확률을 평가하였다. 채점시간을 전체적으로 약 80% 이상 줄일 수 있었고 채점 정확률은 사회 및 국어 과목에 대해 각각 98.7%와 97.2%로 나타났다. 앞으로 자동 채점 시스템의 성능을 개선하고 인간 채점자의 집중도를 높일 수 있도록 인터페이스를 개선한다면 국가수준의 대단위 평가에 충분히 활용할 수 있을 것으로 생각한다.

• Journal of Science Education
• /
• v.41 no.3
• /
• pp.382-404
• /
• 2017

This study aims to explore a learning progression for eight core concepts of middle school science using constructed response items in the national assessment of educational achievement (NAEA). For this purpose, a total of 7,390 responses of the 9th graders on 8 constructed response items in NAEA conducted in 2015 were inputted by computer and the inputted answers were analyzed and classified into different levels. After completing this process, five levels were set for each construct and a construct map was created according to the levels. In this study, the construct map was written in a progression-based rubric format, which was used as a criterion for leveling the answers of all students. In addition, the Rasch model was applied to measure the scores and achievements of the respondents by means of statistical analysis to correlate the scores of the students with the achievement levels of the constructs. Lastly, a preliminary learning progression was created by revising the construct map reflecting the results of Rasch model application. This study was meaningful in that it explored the possibility of developing the learning process by using constructed response items in NAEA. However, the preliminary learning progression developed in this study is still hypothetical and inferential because it is not longitudinally traced to individual students. Therefore, it is necessary to continually revise and supplement through iterative research process.

• KIPS Transactions on Software and Data Engineering
• /
• v.5 no.11
• /
• pp.527-534
• /
• 2016

The emergent information society requires the talent for creative thinking based on problem-solving skills and comprehensive thinking rather than simple memorization. Therefore, the Korean curriculum has also changed into the direction of the creative thinking through increasing short-answer questions that can determine the overall thinking of the students. However, their scoring results are a little bit inconsistency because scoring short-answer questions depends on the subjective scoring of human raters. In order to alleviate this point, an automated scoring system using a machine learning has been used as a scoring tool in overseas. Linguistically, Korean and English is totally different in the structure of the sentences. Thus, the automated scoring system used in English cannot be applied to Korean. In this paper, we introduce an automated scoring system for Korean short-answer questions using predictability and unanimity. We also verify the practicality of the automatic scoring system through the correlation coefficient between the results of the automated scoring system and those of human raters. In the experiment of this paper, the proposed system is evaluated for constructed-response items of Korean language, social studies, and science in the National Assessment of Educational Achievement. The analysis was used Pearson correlation coefficients and Kappa coefficient. Results of the experiment had showed a strong positive correlation with all the correlation coefficients at 0.7 or higher. Thus, the scoring results of the proposed scoring system are similar to those of human raters. Therefore, the automated scoring system should be found to be useful as a scoring tool.

• Journal of the Korean earth science society
• /
• v.39 no.1
• /
• pp.103-116
• /
• 2018

The purpose of this study was to explore $6^{th}$ graders learning progression for lunar phase change focusing astronomical systems thinking. By analyzing the results of previous studies, we developed the constructed-response items, set up the hypothetical learning progressions, and developed the item analysis framework based on the hypothetical learning progressions. Before and after the instruction on the lunar phase change, we collected test data using the constructed-response items. The results of the assessment were used to validate the hypothetical learning progression. Through this, we were able to explore the learning progression of the earth-moon system in a bottom-up. As a result of the study, elementary students seemed to have difficulty in the transformation between the earth-based perspective and the space-based perspective. In addition, based on the elementary school students' learning progression on lunar phase change, we concluded that the concept of the lunar phase change was a bit difficult for elementary students to learn in elementary science curriculum.

• Annual Conference on Human and Language Technology
• /
• 2012.10a
• /
• pp.69-72
• /
• 2012

본 논문은 한국어 서답형(단어, 구 수준) 문항 유형을 분석하고 실제 채점자가 채점 기준표를 보고 채점하는 방법을 컴퓨터가 인식할 수 있도록 정답 템플릿을 설계 및 개념 정의를 하여 한국어 서답형에 특화된 자동채점 시스템 방법을 제시한다. 본 시스템을 사용하여 1000개의 학생 답안지에 대한 유형 가지수 500개 이하의 2011년도 학업성취도 평가 과학 6개 문항에 대하여 채점 기준표 내용을 정답 템플릿으로 작성한 뒤 250개 학생 답안을 학습데이터로, 정답 템플릿을 업데이트로 사용, 750개 학생 답안에 대하여 자동채점한 결과, 평균 카파계수 0.84라는 수치로서 실제 사람 채점 결과와 거의 완벽히 일치라는 결과를 얻었다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2014.11a
• /
• pp.909-912
• /
• 2014

최근 자연어 처리 분야에서는 한국어로 작성된 서답형 문항을 자동으로 채점하기 위한 시스템을 개발하기 위한 연구가 꾸준히 진행되고 있다. 2013년도에는 답안 정규화, 채점 기준 작성 및 편집, 자동 채점 및 채점 후처리 작업 등 자동 채점 프로세스를 수행할 수 있는 시스템을 설계 및 개발하였다. 본 논문은 2013년도에 개발된 자동 채점 시스템의 문제점을 분석하고 채점자가 보다 편리하게 채점을 수행할 수 있도록 추가 및 개선한 기능과 이를 구현하기 위해 이용한 자연어 처리 기술을 소개한다.

• Korean Journal of Cognitive Science
• /
• v.26 no.2
• /
• pp.147-165
• /
• 2015

Through short-answer questions, we can reflect the depth of students' understanding and higher-order thinking skills. Scoring for short-answer questions may take long time and may be an issue on consistency of grading. To alleviate such the suffering, automated scoring systems are widely used in Europe and America, but are in the initial stage in research in Korea. In this paper, we propose a semi-automatic scoring system for short Korean free-text responses using semi-supervised learning. First of all, based on the similarity score between students' answers and model answers, the proposed system grades students' answers and the scored answers with high reliability have been included in the model answers through the thorough test. This process repeats until all answers are scored. The proposed system is used experimentally in Korean and social studies in Nationwide Scholastic Achievement Test. We have confirmed that the processing time and the consistency of grades are promisingly improved. Using the system, various assessment methods have got to be developed and comparative studies need to be performed before applying to school fields.

• Journal of the Korean Geographical Society
• /
• v.49 no.4
• /
• pp.615-632
• /
• 2014

This study aims to explore 'bluffing', one of the most important topics in order to ensure the objectivity, validity, reliability of scoring of constructed-response items. The author identifies the conception of bluffing, and classifies major types of bluffing on the basis of previous studies on the theoretical level. Next, the author analyzes empirically the bluffing strategies and types of learners on key terms of Korean Geography subject. Compared with the existing research reports, the result of this study shows a significantly lower average bluffing score. On the other hand, it is consistent in results of previous studies reported that average bluffing score is similar between genders and that those students who got highest grades did least bluffing. Actually bluffing types are classified into four categories: 'repeating the question' type, 'depending on a well-known or existing knowledge' type, 'applying some basic concepts regardless of understanding' type, and 'inducing scorer's basic beliefs' type. Some comments and suggestions are as follows. First, it is necessary to continue researches of the relations among bluffing and age, gender, grade, intelligence and learning styles. Second, it is required for scorers who score constructed-response items to develop and supply the scoring guide including analysis contents of bluffing types and cases, and increase opportunity for training. Third, we need to inquire the domain-specific bluffing types in geography subject based on the generalizable sample size.

• Journal of The Korean Association For Science Education
• /
• v.31 no.8
• /
• pp.1158-1174
• /
• 2011

In this study, we examined the compositions, basic principles, and the area of the National Higher Education Entrance Examination (GaoKao) in 2009, we also analyzed the categories and characteristics of items. Also, the GaoKao was analyzed in terms of test specifications, the number of items, item patterns, difficulty levels, and implications of the College Scholastic Ability Test(CSAT) were explored. Results show Natural Science section of the National Test 1, 2 are 300 points per 150 min, and Natural Science, and Chemistry of Shanghai is 150 points each per 120 min. Also, the GaoKao contained multiple choice and fill in the blanks questions, and the description items are composed of experiments of various types. The GaoKao Natural Science section is composed of physics, chemistry, biology but not earth science, which is different from the CSAT. GaoKao requires basic understanding or the observation ability to reasoning, the complex thinking ability, especially emphasized on the experiment ability. The range of possible questions is in the examination outline, not the curriculum, and the ratio of questions from the University level is high. In the analysis of the behavioral domain, the ratios of the understanding and application items is higher than the CSAT, and inquiry items is lower, but the inquiry items are deeper. In case of the ratio of the expected correct answer, National Test 1 and National Test 2 is similar, but the difficult items or about 20~39% of the test is 4~5 times to that of the CSAT, making the GaoKao very difficult. The peculiar characteristics of GaoKao is the emphasis on the experiment, and even though the practical items is of lower ratio, they are very useful in life.

• Communications of Mathematical Education
• /
• v.26 no.1
• /
• pp.29-50
• /
• 2012

The purpose of this study is a comparison of proficient level and below basic level students' mathematical achievement in the National Achievement Evaluation and Assessment(NAEA). For the purpose, this study compared the proficient level and below basic level students' ratios, students' mathematical achievement of contents area and behavioral area in 6th, 9th, and 11th grades. This study found the change of proficient level and below basic level students' ratios, and the proficient level and below basic level students' characteristics on mathematical achievement of contents area and behavioral area in 6th, 9th, and 11th grades.