现代远程教育研究
現代遠程教育研究
현대원정교육연구
MODERN DISTANCE EDUCATION RESEARCH
2015年
3期
104-112
,共9页
信息可视化%文本分析%可视化工具%操作方法%案例研究
信息可視化%文本分析%可視化工具%操作方法%案例研究
신식가시화%문본분석%가시화공구%조작방법%안례연구
Information Visualization%Text Analytics%Visualization Tools%Operation Method%Case Study
可视化文本分析表达了一种对文本数据运用交互式图形呈现方式,实现知识发现的信息分析技术和过程,其应用过程一般分为文本处理、可视化呈现和交互理解三个阶段。进行文本可视化分析操作时,需根据研究对象的特征,选择恰当的工具,从原始材料中抽取文本的特征属性或元数据,在合适的视觉编码描绘和概括文本内容、结构、关系等基础上,与用户互动,揭示文本信息的特征和规律。已有研究表明,借助技术在计算和可视化上的能力,可视化文本分析技术可以弥补人工分析时存在的耗时长、主观性强等问题,提升文本信息处理与理解的效率,深入探察数据中隐藏的特征、关系和模式。基于全国教育科学规划教育技术类课题的案例研究验证了这些优势,且这一做法正逐渐引发业内的研究关注,成为一大发展趋势。案例研究还发现:受中文自然语言处理技术还不够成熟的影响,可视化文本分析在中文文本应用中还比较有限,在分词、工具选用以及分析深度等方面还存在不足。
可視化文本分析錶達瞭一種對文本數據運用交互式圖形呈現方式,實現知識髮現的信息分析技術和過程,其應用過程一般分為文本處理、可視化呈現和交互理解三箇階段。進行文本可視化分析操作時,需根據研究對象的特徵,選擇恰噹的工具,從原始材料中抽取文本的特徵屬性或元數據,在閤適的視覺編碼描繪和概括文本內容、結構、關繫等基礎上,與用戶互動,揭示文本信息的特徵和規律。已有研究錶明,藉助技術在計算和可視化上的能力,可視化文本分析技術可以瀰補人工分析時存在的耗時長、主觀性彊等問題,提升文本信息處理與理解的效率,深入探察數據中隱藏的特徵、關繫和模式。基于全國教育科學規劃教育技術類課題的案例研究驗證瞭這些優勢,且這一做法正逐漸引髮業內的研究關註,成為一大髮展趨勢。案例研究還髮現:受中文自然語言處理技術還不夠成熟的影響,可視化文本分析在中文文本應用中還比較有限,在分詞、工具選用以及分析深度等方麵還存在不足。
가시화문본분석표체료일충대문본수거운용교호식도형정현방식,실현지식발현적신식분석기술화과정,기응용과정일반분위문본처리、가시화정현화교호리해삼개계단。진행문본가시화분석조작시,수근거연구대상적특정,선택흡당적공구,종원시재료중추취문본적특정속성혹원수거,재합괄적시각편마묘회화개괄문본내용、결구、관계등기출상,여용호호동,게시문본신식적특정화규률。이유연구표명,차조기술재계산화가시화상적능력,가시화문본분석기술가이미보인공분석시존재적모시장、주관성강등문제,제승문본신식처리여리해적효솔,심입탐찰수거중은장적특정、관계화모식。기우전국교육과학규화교육기술류과제적안례연구험증료저사우세,차저일주법정축점인발업내적연구관주,성위일대발전추세。안례연구환발현:수중문자연어언처리기술환불구성숙적영향,가시화문본분석재중문문본응용중환비교유한,재분사、공구선용이급분석심도등방면환존재불족。
Visual text analytics shows a kind of information analysis techniques and processes of using interactive graphical methods to achieve knowledge discovery. Its application has three steps, including text processing, visual presentation and interactive interpretation. First, feature attributes or metadata should be extracted from the original text material with appropriate tools according to the characteristics of the study object. Then, based on the proper visual coding to describe and summarize the content, structure and relations of texts, traits and rules of the textual information are discovered through user interaction. Studies have shown that, with the capability of calculation and visualization of the technology, visual text analytics can make up the problems existing in manual analysis such as the time-consuming and subjectivity, enhance the efficiency of text information processing and understanding, and deeply explore the hidden characteristics, relationships and patterns of data. A case study carried out on the topics of educational technology research projects of the National Education Science Plan from 2006 to 2013 verified the advantages of visual text analytics in text comprehension. Visual text analytics is gradually arousing research attention and becomes a major trend. The case study also discussed the weakness of title tokenization, the inadequacies of research tools to support Chinese text analytics and the insufficient utilization of text features. Affected by the immaturity of Chinese natural language processing technology, the application of visual text analytics in Chinese is still limited.