초록

질의응답 시스템에서의 질의 분석 과정은 이용자의 자연어 질의 문장에서 질의 의도를 파악하여 그 유형을 분류하고 정답 추출을 위한 정보를 구하는 것이다. 본 연구에서는 복잡한 분류 규칙 집합이나 대용량의 언어 지식 자원 대신 이용자 질의 문장에서 질의 초점 어휘를 추출하고 구문 구조적으로 관련된 단어들의 의미 정보에 기반하여 효율적으로 질의 유형을 분류하는 방법을 제안한다. 질의 초점 어휘가 생략된 경우의 처리와 동의어와 접미사 정보를 이용하여 질의 유형 분류 성능을 향상시킬 수 있는 방법도 제안한다.

Abstract

For question-answering system, question analysis module finds the question points from user’s natural language questions, classifies the question types, and extracts some useful information for answer. This paper proposes a question type classifying technique based on focus words extracted from questions and word semantic information, instead of complicated rules or huge knowledge resources. It also shows how to find the question type without focus words, and how useful the synonym or postfix information to enhance the performance of classifying module.

초록

This experimental study proposes a multi-document summarization method that produces optimal summaries in which users can find answers to their queries. In order to identify the most effective method for this purpose, the performance of the three summarization methods were compared. The investigated methods are sentence clustering, passage extraction through spreading activation, and clustering-passage extraction hybrid methods. The effectiveness of each summarizing method was evaluated by two criteria used to measure the accuracy and the redundancy of a summary. The passage extraction method using the sequential bnb search algorithm proved to be most effective in summarizing multiple documents with regard to summarization precision. This study proposes the passage extraction method as the optimal multi-document summarization method. 攀＊＊＊ 본 연구는 연세대학교 대학원 박사학위논문의 일부를 요약한 것임.＊＊＊ 연세대학교 문헌정보학과 시간강사(shchoi@lis.yonsei.ac.kr)＊＊＊＊연세대학교 문헌정보학과 교수(ymchung@yonsei.ac.kr) 논문접수일자 : 2004년 8월 27일 게재확정일자 : 2004년 9월 13일攀攀

Abstract

초록

대용량 문서에서 정보를 추출하는 작업은 정보검색 분야뿐 아니라 질의응답과 요약 분야에서 매우 유용하다. 정보추출은 비정형 데이터로부터 정형화된 정보를 자동으로 추출하는 작업으로서 개체명 인식, 전문용어 인식, 대용어 참조해소, 관계 추출 작업 등으로 구성된다. 이들 각각의 기술들은 지금까지 독립적으로 연구되어왔기 때문에, 구조적으로 상이한 입출력 방식을 가지며, 하부모듈인 언어처리 엔진들은 특성에 따라 개발 환경이 매우 다양하여 통합 활용이 어렵다. 과학기술문헌의 경우 개체명과 전문용어가 혼재되어 있는 형태로 구성된 문서가 많으므로, 기존의 연구결과를 이용하여 접근한다면 결과물 통합과정의 불편함과 처리속도에 많은 제약이 따른다. 본 연구에서는 과학기술문헌을 분석하여 개체명과 전문용어를 통합 추출할 수 있는 기반 프레임워크를 개발한다. 이를 위하여, 문장자동분리, 품사태깅, 기저구인식 등과 같은 기반 언어 분석 모듈은 물론 이를 활용한 개체명 인식기, 전문용어 인식기를 개발하고 이들을 하나의 플랫폼으로 통합한 과학기술 핵심개체 인식 체계를 제안한다.

Abstract

Large-scaled information extraction plays an important role in advanced information retrieval as well as question answering and summarization. Information extraction can be defined as a process of converting unstructured documents into formalized, tabular information, which consists of named-entity recognition, terminology extraction, coreference resolution and relation extraction. Since all the elementary technologies have been studied independently so far, it is not trivial to integrate all the necessary processes of information extraction due to the diversity of their input/output formation approaches and operating environments. As a result, it is difficult to handle scientific documents to extract both named-entities and technical terms at once. In order to extract these entities automatically from scientific documents at once, we developed a framework for scientific core entity extraction which embraces all the pivotal language processors, named-entity recognizer and terminology extractor.

초록

Abstract

Although it has been a long subject of study why researchers prefer some cited documents to others, the existing relative researches have had a variety of perspectives on the nature and complexity of the citation behavior and not provided a complete answer to this question. In particular, Korea researchers mainly used statistical analysis of bibliographic information, which has limitations in revealing dynamic and complex cognitive aspects of the citation process. In this study, I investigate the citer perception of citing motives and bibliographic factors through survey and compared the responses according to the researchers’ characteristics. After extracting the 22 motivations and 21 factors through the literature analysis and configuring a 5-point Likert scale questions, I conducted a survey in the wat of an e-mail attachment. From the SPSS 22.0, the frequency analysis, t-test, and one-way ANOVA were performed on the 354 valid samples. As a result, it is found that supporting is considered the most important citing motive and social connection, self-citation have little influence. In the case of bibliographic factors, the journal’s reputation was recognized the most influential factor and the number of pages and authors was the least. Significant differences in fields of study and research careers were showed in some parts. These results can substantiate earlier studies, determine whether the factors assumed influential in selecting references were intended, and suggest the search point to the specialty library or academic database.