Application of Nominal Indirect Anaphora in English Teaching and Learning

This paper deals with the application of nominal indirect anaphora to unveil the essence of nominal indirect anaphora and disclose the magic power of cohesion and coherence inside it from the angle of cognitive linguistics.Thus we can apply the resolution mechanism for English teaching and learning for English learners.

Anaphora＇s Creating Irony by the Aid of Context

The core function of anaphora is to serve as a substitute for a certain linguistic component in texts, but it actually has other missions such as creating the rhetorical effect--irony. However, anaphora alone cannot contribute to ironic effect, and only by resorting to other resources like contextual information, can anaphora generate irony. Therefore, Yus＇ s classification of contextual sources is employed to analyze with examples how anaphora creates irony by the aid of context. Ultimately, the conclusion has been drawn that anaphora is able to produce irony with the help of six groups of contextual information classified by Yus except the category--speaker＇s nonverbal behavior and anapbora sometimes has to interact with several groups of contextual information in order to create irony.

Linguistic Context's Effect to Anaphora

The anaphora resolution firstly relies on the special association between the anaphor and the antecedent.But this kind of associa tion may not result in the accomplishment of the anaphora resolution.By the analysis of many kinds of anaphoric examples,this paper will point out that,when the anaphora resolution has some mistake,the linguistic context will restrict on anaphor.The purpose is to influence semantic connection between the antecedent and the anaphor,and make the anaphora resolution accomplish successfully.

Mental Space Motivation for Anaphora and Exphora

Mental space theory(MST),as an important theory in cognitive linguistics is a universal cognitive theory. And it can help greatly in explaining the dynamics,random and vague cognitive process. This paper tries to use this theory to explain how direct and indirect anaphora and exphora work in a discourse.

Aspects of anaphora in Chinese and in some Germanic,Romance,and Slavic languages,the‘syntactic'versus‘pragmatic'language typology,and neo-Gricean pragmatics

In this article,two types of anaphora—(i)null subjects and(ii)long-distance reflexivisation— will be briefly compared and contrasted between Chinese and some Germanic,Romance,and Slavic languages,showing how they are different typologically.Following Huang(2000a,2013a),utilising intra-sentential anaphora as a testing ground,I shall re-hypothesise that languages in the world can roughly be divided into two groups:syntactic and pragmatic.Finally,I shall outline an analysis of long-distance reflexivisation in Chinese in terms of my neo-Gricean pragmatic theory of anaphora.

How Language Aims at a Target:The cognitive system underlying deixis and anaphora

As a speaker communicates with a hearer,the speaker’s attention can come to be on a particular entityher " target"—that she wants to communicate about to the hearer.This target can be located near or far in either the speech-external(deictic) or the speech-internal(anaphoric) environment.She thus needs the hearer to know what her intended target is and to have his attention on it jointly with her own.The problem,though,is how to bring this about.She cannot somehow reach into the hearer’s cognition and directly place his focus of attention on her selected target.Language solves this problem by initiating a certain two-stage procedure in the hearer’s cognition.For this function,it uses a specialized set of mostly closed-class lexical forms;" triggers".English triggers include;this/ these,that/those,here,there,yonder,now,then,thus,so,such,yay,thisaway,thataway,personal pronouns, relative pronouns,and tense markers.The speaker initiates the hearer’s procedure by placing a trigger at the relevant point in the current sentence of her discourse.In the first stage,the trigger directs the hearer to find certain elements of information to which he does have ready access.These are "cues" to the speaker’s intended target.Such cues belong to ten distinct categories,representing ten different sources of information.In the second stage,equipped with the cues he has found,the hearer uses them in combination to determine the speaker’s intended target.

汉语说明语篇中的事物回指研究

本文以汉语说明语篇中与说明对象相关的回指形式为研究对象,通过定量统计和定性分析相结合的方法,考察各种不同类型回指形式在说明语篇中的篇章分布情况,归纳说明语篇中有关说明对象的回指类型。分析汉语说明语篇中的回指在编码时遵循的倾向性规律和一般性原则,并进一步探讨回指作为一种特殊语篇资源的多种功能和特殊价值。

从“数量名+NP”话题句看话题和述题的复杂依存关系

汉语“数量名+NP”在特定语境中可作话题,被赋予定指理解,但同时也凸显数量特征,是有标记话题。该结构要求述题须表达数量特征,但受句子体特征和语用因素影响其内部也有细微差别。“数量名”话题自身的特征与“话题—述题”的陈述关系决定结构对述题的要求,而述题的构造和语义也有助于话题指称类型的理解,二者相互依存。

语言的界限:“情境论”视角下的布兰顿回指理论

通过回指理论提供的以“次语句”为核心的推论网络,布兰顿将“语言-世界”的关系转变为“语言-语言”的关系,为刻画指称关系提供了新进路。但是,布兰顿的回指理论面临两个主要困难:“指称不透明”和“回指不共指”。为了解决这两个问题,将采用雷卡纳蒂的“情境论”来优化回指理论,通过引入σ-异质结构来解决布兰顿回指理论面临的困难。首先,在回指链中插入情境算子以揭示句子的情境结构,严格区分句子的情境内容和语义内容;其次,厘清不同认知主体在不同情境中的回指链及其逻辑关系;最后,那些位于回指链之外无法穷尽的“最终视角”和无法还原的现实世界,同样以情境方式呈现出来。总之,在回指关系中“语言”“心灵”和“世界”都是必不可少的一环。

主语属格语的生命性及其相关问题

主语属格语充当零形回指先行语时,主语属格语结构在生命性特征分布上缺失“无生—有生”这一组合,构成一种蕴涵关系。同时,数据分布及理论分析显示,回指研究的可及性等级排序中,主语属格语后置于主语更符合语言事实。主语属格语之所以能够充当回指先行语与生命性、语法位置、主题化能力及语言的基本组织方式等因素有关,生命性只是其中一个比较容易感知到的语义因素,在具体的语言环境中,主语属格语的高可及性是多种因素综合作用的结果。

动态指称与汉语话题结构允准

认知语用上,汉语话题结构的允准机制,可以解读为话题对述题的指示、指别以及述题对话题的语义回指等动态指称内容。首先,话题作为信息处理优化器(PEMs),通过所指实体、时间、空间、原因、条件、领属、范围等信息来限定述题所述的适用范围,相当于对述题内容的指别;其次,在信息内容上,为了充分满足述题信息的完整性,述题必然存在语义空位与话题成分复指或回指。语义空位在述题中表现为但不限于复指代词、附缀、句法空位甚至是附加语等句法形式。

Similarities and Differences of Anaphoric Pronoun in Chinese and English

In both English and Chinese, pronouns are used with great frequency. And there are numerous usages of pronoun anaphors. The similarities and differences are summarized with the comparison of pronoun anaphors in English and Chinese.

非结构化信息抽取关键技术研究探讨

以基于内在认知机理的知识发现理论为指导,针对汉语命名实体识别的难点,充分考虑专家知识在命名实体识别中的作用;根据不同的实体类型,采用灵活变化的统计与规则相结合的方式;采用各种技术来研究信息抽取的任务,如:机器学习技术、篇章分析与理解技术、句法分析技术、图算法与图挖掘技术、词计算技术、快速全文检索技术等;该文探讨的是不仅要从文本中获取简单子句中的关系,还要获得跨句子、段落中的实体关系。

文本阅读中空间距离的心理表征

采用学习探测模式和回指解决相结合的研究范式,探讨情境模型中空间距离的心理表征问题。实验1分离几何距离和类别距离,探讨类别距离对回指解决的影响。结果表明,在控制几何距离的条件下,读者采用由房间数目体现的类别距离信息建构情境模型;实验2进一步探讨几何距离的作用,结果发现,类别距离信息相同的条件下,读者在情境模型中表征几何距离信息。研究表明,类别距离和几何距离分别对空间情境模型的回指解决产生独立影响。

信息抽取研究综述

信息抽取的任务是从大量数据中准确、快速地获取目标信息,提高信息的利用率。目前,信息抽取已经成为NLP领域的一个重要分支。随着互联网应用的发展,其价值也正日益显现,学术界和工业界对此都寄予厚望。首先回顾了信息抽取的发展历程;接着从命名实体识别、指代消解、关系抽取和事件抽取4个方面总结了信息抽取关键技术的研究进展;然后分析了信息抽取目前面临的若干主要问题;最后对信息抽取的研究趋势作了预测。

“自己”是谁?——对约束原则的实验研究

本文通过两个行为实验,考察了无语境条件下,汉语反身代词“自己”回指的认知加工方式。实验一采用语句选择(sentencegating)作业,考察了被试利用韵律信息确定反身代词“自己”回指的可能性,结果表明听着很难利用韵律信息确定反身代词“自己”的回指;实验二利用跨通道启动(crossmodalpriming)和词汇命名作业,考察了被试对反身代词“自己”的认知加工,结果表明当目标词和小主语有关时,反身代词“自己”对目标项的启动效应要大于目标词和大主语有关,表现出“局部约束”(localbinding),这与约束理论的A原则是符合的,而与汉语反身代词的“长距离回指”现象相矛盾。

指代消解中语义角色特征的研究

该文实现了一个基于机器学习的指代消解平台,并在此基础上着重研究了语义角色特征对指代消解的影响。该文使用ASSERT①语义角色标注系统得到语义角色标注信息,然后在原型系统的基础上加入语义角色特征。为了分析语义角色特征对指代消解的影响,该文还分析了语义角色特征和指代链特征以及代词细化特征的结合对系统的影响。通过把先行语和照应语在句子中所作的语义角色特征加入机器学习系统中进行研究,该文发现语义角色特征能够显著提高系统的性能,特别是对代词的消解有很好的效果。在ACE 2003 NWIRE基准语料上的所有类型名词短语的指代消解测试表明,召回率提高了3.4%,F值提高了1.8%。

句子重述技术在中文文本摘要中的应用

本文提出了一种对中文文本摘要中抽取出的句子进行重述的方法。首先使用基于统计的方法对文本进行特征统计，计算词和句子的权重，摘取出权值较高的句子；然后对这些句子应用一种基于向量相似度计算的算法进行指代消解，同时提出一种新的句子向量相似度的计算方法去除冗余；最后利用启发式规则进行加工，从而得到文本摘要。实验结果显示，系统修改后的文摘具有较好的连贯性和流畅性，与修改之前的文摘相比，文摘质量有明显提高。

汉语中指代问题的研究及讨论

指代的解决对于自然语言理解的非歧义具有重要的意义重视篇章和故事情节的理解固然有助于解决类似的二义性问题，但所需的巨大知识库在实际系统中，无论是建立还是最终的效率和完备性均很难满足用户的需要．文章以第三人称代词为代表，分析其各种指代现象并进行分类通过语法分析和句子语义平面的主客观信息研究，利用格框架，提出了在上下文相关的语义环境中进行指代分类解决的思想，并给出了相应算法．

采用优先选择策略的中文人称代词的指代消解

指代是自然语言中常见的语言现象,指代消解是文本信息处理中的一个重要任务。随着篇章处理相关应用日益广泛,指代消解也显示出前所未有的重要性。本文针对中文人称代词的指代特点,提出了一种基于语料库的,运用决策树机器学习算法并结合优先选择策略,进行指代消解的方法。该方法充分考虑了与指代相关的若干属性,及相互之间的影响。实验表明,对中文人称代词的消解特别是第三人称的消解获得了一定的效果。