自然语言理解中的语义分析问题

自然语言理解基于语义理解，尤其是汉语，努力挖掘有形式依据的深层次的语义组织，进行深入的语义研究，成为21世纪自然语言理解的突破口。从计算机模拟人类使用自然语言的方式，汉语重意合的特点说明了加强语义分析的必要性，又从计算机理解自然语言的过程举例说明了语义分析的几个问题。     

基于表征重述的机器文本理解

传统机器翻译系统缺乏联系上下文形成认知的能力,仅根据对应单词的默认含义进行翻译,容易导致语义错误等问题。通过模拟人的表征重述认知过程,提出一种新的机器文本理解与翻译方法。该方法可通过较少的实例对文本进行理解和翻译,避免出现语义理解错误问题,且无需进行繁杂的语法标注。实验表明,该方法可通过引入习得的常识,使翻译出现歧义错误的概率降低到 1%以下,并可标注出不符合常理而又无法找到更好解释的句子。     

听力理解过程中语义联想的误区及其对策

指出语义联想是听力理解过程中一项重要技能,如何指导学生进行正确的语义联想是提高英语听力理解的重要途径之一。从词汇、语境、文化背景等方面入手,分析造成语义联想偏差的具体原因,并提出解决这一问题的方法和途径。     

自然语言理解中的语义分析问题

自然语言理解基于语义理解 ,尤其是汉语 ,努力挖掘有形式依据的深层次的语义组织 ,进行深入的语义研究 ,成为 2 1世纪自然语言理解的突破口。从计算机模拟人类使用自然语言的方式、汉语重意合的特点说明了加强语义分析的必要性 ,又从计算机理解自然语言的过程举例说明了语义分析的几个问题     

“双元结构”下的概念转换

由于中国成年人在学习外语时有“双元结构”的存在,因而造成了成年人在进行语义理解时对不同语言的不同原型不能理解和转换,这时就需要借助概念进行转换,而在概念转换时激活不同文化模式下认知域,从而达到匹配语义、理解语义的目的,当不能匹配时,就需文化的介入,进而完成概念转换。     

文化语境对交际话语语义理解的影响

文化语境是语境中影响交际话语语义理解的主要因素。英汉两种文化在长期的历史发展过程中形成了各具特色的社会习俗、社会规范和价值观念,可以说不同的语言就是不同文化的反映。汉、英两种语言中都带有鲜明的文化特征,在交际的过程中,文化语境对跨文化交际的双方的话语理解起着至关重要的作用。正确把握文化语境对交际话语语义理解的影响是实现成功的英汉交际的前提条件和重要保障。     

阅读理解与英语四级考试

阅读理解是四有考试中的关键部分，如何提高阅读理解的能力一直是教师和学生关注的问题。本文作者试图从对语篇长难句的理解，生词语义的判断力，题意的理解等几个方面的问题来分析影响学生阅读理解能力的因素，以探求更有效的教学方法。     

分数概念语义理解对儿童乘法应用题表征的影响

理解分数概念语义含义,意味着能用分数符号表示不同问题情境中两个量的关系.而分数应用题表征的关键,是将问题情境中事实关系正确转换为分数运算.二者关系尚缺乏实证研究加以检验.以295名小学六年级学生为被试,考察了儿童分数概念语义理解对乘法应用题表征的影响.结果表明:(1)儿童分数概念语义理解可以整体预测乘法应用题表征水平;(2)儿童在部分整体含义和测量含义的理解水平,均可单独预测其对计量、比较和转换乘法应用题表征水平;(3)儿童对"比"的理解,可单独预测比较应用题表征水平,对"商"的理解,可单独预期转换应用题表征水平,同时,算子含义不能单独预测任何应用题表征水平.这一结果说明,儿童分数概念语义理解影响其表征应用题中的事实关系.     

文本语义表征的认知心理学模型建构——基于"情景语境对语义理解的影响"

在文本理解的过程中,读者不仅对文本信息进行加工形成文本的心理表征,而且对文本意义会形成相应的语义表征。然而语义的理解离不开语境,成功的语义理解基于对语境的正确分析。因此,本文依据情景语境理论,举例分析了情景语境对语义理解的影响;并在此基础上探讨了文本理解中语义表征形成的认知心理过程,进而构建了文本理解中语义表征的认知心理学模型。     

英语考试中阅读理解常见题型解析

阅读理解历来是各类英语考试中的重点和难点,由于其题型灵活、覆盖面广,所以许多学生发怵。但是阅读理解也有一定的规律,本文就历年考试中常见题型进行分析总结,希望对英语学习方面有困难的学生有所帮助。一、词汇和语义理解题词汇和语义理解题对词汇理解的考查是阅读测试的一个重要方面。词汇题往往要求对文章中的某个单词、短语甚至句子等找出近义词或最合适的解释。     

基于语义技术的Web服务组合方法研究

针对近几年在语义web服务自动组合上的研究成果,深入归纳和总结了目前语义web服务组合的定义、组合方法以及正确性验证技术。对语义web服务组合问题进行分类和形式化定义,并归纳出语义Web服务自动组合框架。按照不同问题分析和比较相关的组合方法,同时总结了组合方案正确性验证技术。最后,根据前面的分析和总结,提出了语义Web服务自动组合当前面临的问题和挑战。     

Connaissances utilisées par des elèves de 8 à 12 ans dans la formulation de problèmes arithmétiques concrets

This study analyses children development of semantic, linguistic, procedural and schematic knowledge in the context of writing arithmetic word problems. 139 children aged between 8 and 12 years old were presented with a task which consisted in writing arithmetic word problems, according to some contraints: words, questions or measures to include in their problems; type of problems to write. Results show the relevance of actual theoritical models of problem solving (Mayer, 1983; Kintsch & Greeno, 1985). Schematic knowledge seem indeed more important than other knowledge in the process of writing arithmetic word problems; semantic knowledge are also used to choose relevant numbers or measures; the roles of linguistic and procedural knowledge seem less evident. Finally, some hypotheses related with the development of mental models of arithmetic word problems are formulated.     

重庆市万州区农村移民安置超规划有关问题的调查与思考

本文在评价万州区农村移民经验的基础上,重点对该库区农村移民超规划的问题进行了调查剖析,其目的是揭示和分析农村移民安置规划实施10年来,在执行过程中存在的主要矛盾和主要问题,并根据有关移民政策的连续性和问题突出存在的客观性,提出遗留问题解决的措施和政策性建议。     

句法、语义成分的认知语境透视

句法和语义成分是语法研究的两个不可忽视的方面,研究者对它们之间的关系已经趋于一致:语义结构决定句法结构。然而,对于语义结构的来源这一基本问题尚未得到充分研究。文章通过英汉两种语言中的几个个案,试图对语义与句法的相关问题作进一步阐释。结合个案,文章认为语义结构是由认知语境决定的;同时也指出:认知语境对复杂多变的句子形式的解释是不充分的,相关语言现象的解决仍须借助于其它理论。     

What underlies successful word problem solving? A path analysis in sixth grade students

Two component skills are thought to be necessary for successful word problem solving: (1) the production of visual-schematic representations and (2) the derivation of the correct relations between the solution-relevant elements from the text base. The first component skill is grounded in the visual–spatial domain, and presumed to be influenced by spatial ability, whereas the latter is seated in the linguistic–semantic domain, and presumed to be influenced by reading comprehension. These component skills as well as their underlying basic abilities are examined in 128 sixth grade students through path analysis. The results of the path analysis showed that both component skills and their underlying basic abilities explained 49% of students’ word problem solving performance. Furthermore, spatial ability and reading comprehension both had a direct and an indirect relation (via the component skills) with word problem solving performance. These results contribute to the development of instruction methods that help students using these components while solving word problems.     

What is the unknown? The ability to identify the semantic role of the unknown from word problems longitudinally predicts mathematical problem solving performance

The current study aimed to investigate children’s difficulties in word problem solving through assessing their ability to mathematize, or to identify the semantic role of the unknown from word problems. Fifth graders (n = 213) were given an advanced word problem reasoning task in which they had to match word problems with schematic diagrams that depict different processes (multiplication versus division) and the unknown being in different semantic roles (e.g., unit size, number of units, or total in an equal group problem). They were also tested on their mathematical problem solving as well as some potential confounding variables (i.e., intelligence and working memory) and mediators. The ability to identify the semantic role of the unknown was shown to be longitudinally predictive of children’s mathematical problem solving performance even after controlling for the effects of covariates and autoregressor. Such a relation was partially mediated by children’s ability to convert word problems into the correct number sentences/equations. The findings not only highlight the importance of unknown identification in mathematical problem solving process, but also provide a practical tool to assess such an ability.     

Elaborating the structures of a science discipline to improve problem-solving instruction: An account of Classical Genetics' theory structure,function, and development

Situating the conceptual knowledge of a science discipline in the context of its use in the solving of problems allows students the opportunity to develop: a highly structured and functional understanding of the conceptual structure of the discipline; general and discipline-specific problem-solving strategies and heuristics; and insight into the nature of science as an intellectual activity. In order realize these potential learning outcomes, the reconstructions of scientific theories used in problem solving must provide a detailed account of (1) realistic scientific problems and their solutions; (2) problem-solving strategies and patterns of reasoning of disciplinary experts; (3) the various ways that theories function for both disciplinary experts and students; and (4) the way theories, as solutions to realistic scientific problems, develop over time. The purpose of this paper, therefore, is to provide further specificity regarding a philosophical reconstruction of the structure of Classical Genetics Theory that can facilitate problem-solving instruction. We analyze syntactic, semantic and problem-based accounts of theory structure with respect to the above criteria and develop a reconstruction that incorporates elements from the latter two. We then describe how that reconstruction can facilitate realistic problem solving on the part of students.     

基于语义web的网络化制造知识检索系统

针对传统知识检索方法缺少语义层次互操作的技术缺陷,建立了一个基于语义web的制造知识检索系统的体系结构.该结构包括一系列支持知识共享以促进网络化制造协同的机制.提出了一种基于五元组的语义信息检索模型,建立了一套语义网络化制造知识检索测试方法.最后,经过实例验证该系统能显著提高网络化制造模式下制造知识信息检索的准确率和查全率,为解决区域性网络化制造中的相关知识检索问题提供了一条切实可行的途径.     

例说等价无穷小在求函数极限中的应用及推广

文章通过对实例的分析,提出了运用等价无穷小求函数极限的特殊情形,说明了等价无穷小所涉及的题型广泛,合理应用能简化计算,是求函数极限中一种非常普遍、非常快捷的方法。     

基于应用题解题障碍的认知工具模式设计

通过整理近年来关于小学数学应用题的相关研究,总结学生在应用题解题过程中普遍存在的认知障碍,结合现有解题系统分析研究各类型的认知工具对小学应用题解题障碍的支持,为帮助学生选择合适的认知工具提供借鉴,并在此基础上提出基于问题解决障碍的小学数学应用题解题系统的模型构建,为新的认知工具的设计开发提供新的角度。