百度
360搜索
搜狗搜索

evaluation有几个音节,Evaluationi,væljʊ'eʃən 这个音标里逗号和上引号分别代表什么?请英语大神解答~详细介绍

本文目录一览: estimate [?est??meit] ,evaluate [i?v?ljueit]两个重音地方为什么不一样?

【单引号、括号中是词缀或词根的含义】
1、estimate用作动词读作['est?me?t]用作名词读作['est?mit]
其名词形式为estimation[est?'me??(?)n]
形容词形式为estimable['est?m?b(?)l]adj. 可估计的;可尊敬的;有价值的
从构词法来讲:estimate=est(存在)+im+ate(做)→对存在的东西做评价
从词源上讲:中世纪英语后期,from Latin asetimat-'determined,appraised', from the verb aestimare. 名词词义在中世纪英语后期为‘intellectual ability, comprehesion’,后来为‘valuing,a valuation’。动词词义在15世纪后期为‘to think well or badly of sb or sth’,后来为‘regard as being, consider to be’。
总之,estim-是主干,并且几百年前就是这么写的,也是这么读的,所以重音还在e上。
2、evaluate vt.[?'v?lj?e?t]
evaluation n.[?,v?lj?'e??n]
value['v?lju?] n. 值;价值;价格;重要性;确切涵义vt. 评价;重视;估价
valuate vt. 对…作估价
从构词法来讲:evaluate=e(出)+valu(value价值,名词词根)+ate(做,动词词缀)评定出价值→评价,估价
从词源上讲:19世纪中叶源自法语evaluer, from es-(from Latin ex-)'out,from出'+Old French value 'value'。
总之,value是词根,是主干,e是为了构造新词加上去的,所以重音还在value上
同理,再举一个例子
evacuate[?'v?kj?e?t] vt./vi. 疏散,撤退;排泄
evacuation[?,v?kj?'e??(?)n] n.疏散,撤退;排泄
vacuum['v?kj??m]n. 真空;空间;真空吸尘器adj. 真空的;利用真空的;产生真空的vt. 用真空吸尘器清扫
vacuate ['v?kj?,et]vt. 在…中造成真空,抽空
evacuate=e+vacu+ate【我就不解释了,重音还在主干vacu-‘vacuum’上面】
鄙人乃英专,长期没看语法词干词根词缀已经傻傻分不清楚,以上可能会有术语用错,还请见谅。估计楼主不是英专,所以也没有查专业书,太多术语看得人头晕,所以就参考了一下词典,构词法参见《新东方 雅思词汇》,词源参见《新牛津英语大辞典》。大学语法书上讲的天花乱坠还有一大堆特例和方言用法,其实不是搞学术研究,真的没必要纠结这个重音规律,我分析的只是根据多年经验的总结,并非教材上的规律,印象当中教材里也没怎么讲过重音还有规律,除了increase动词读作[?n'kri?s]名词读作['?nkri?s]【一个词既可作名词又可作动词,做名词重音在前,做动词重音在后,例子还有decrease等】。多读多听多背,对这个词印象深刻了,重音自然记得。

Evaluationi,v?lj?'e??n 这个音标里逗号和上引号分别代表什么?请英语大神解答~

逗号 是次重音符号
上引号是重音符号
希望能帮到你(*^__^*)
逗号是次重音,引号是重音
下面的应该次重音符号,上面的是重音符号。
逗号代表之后是次重读音节,上引号代表之后是重度音节

评估 英文

评估的英文是evaluation。
评估是一种系统性的过程,用于收集、分析和解释有关某个实体或行为的信息,以便做出判断和提供有关改进措施的建议。评估是一种系统性的过程,通过收集、分析和解释信息,来判断一个实体或行为的质量、有效性和效果。评估的目的是提供可靠的数据和结论,以便做出决策、改进和提升。
评估的类型
评估可分为多种类型,包括学习评估、项目评估、组织评估、政策评估等。每种类型的评估都有其特定的目的和方法,但都旨在提供有关实体或行为的信息和建议。
评估的步骤和方法
评估通常包括以下几个步骤:确定评估目标和问题、设计评估方案、收集和分析数据、制定评估结论和建议,并对评估过程进行反思和改进。评估可以使用定性和定量的方法,如调查问卷、访谈、观察、统计分析等。
评估的重要性和应用领域
评估在各个领域都具有重要的应用价值。在教育领域,评估可以帮助了解学生的学习情况和教学效果,以改进教育质量;在项目管理中,评估可以帮助了解项目的可行性和效果,并提供决策依据;在组织管理中,评估可以帮助了解组织绩效和发展方向,以实现持续改进。
评估的挑战和注意事项
评估也面临一些挑战和注意事项。例如,评估需要确保数据的准确性和可靠性,需要选择合适的评估工具和方法,还需要考虑评估过程中可能出现的偏见和利益冲突等因素。
总结:
评估是一种系统性的过程,通过收集、分析和解释信息,判断一个实体或行为的质量和效果。评估在各个领域都具有重要的应用价值,可以帮助做出决策、改进和提升。然而,评估也面临着一些挑战和注意事项,需要确保数据的准确性和可靠性,并选择合适的评估工具和方法。

NLP基础知识和综述

一种流行的自然语言处理库、自带语料库、具有分类,分词等很多功能,国外使用者居多,类似中文的jieba处理库
为单词序列分配概率的模型就叫做语言模型。
通俗来说, 语言模型就是这样一个模型:对于任意的词序列,它能够计算出这个序列是一句话的概率。或者说语言模型能预测单词序列的下一个词是什么。
** n-gram Language Models **
N-gram模型是一种典型的统计语言模型(Language Model,LM),统计语言模型是一个基于概率的判别模型.统计语言模型把语言(词的序列)看作一个随机事件,并赋予相应的概率来描述其属于某种语言集合的可能性。给定一个词汇集合 V,对于一个由 V 中的词构成的序列S = ?w1, · · · , wT ? ∈ Vn,统计语言模型赋予这个序列一个概率P(S),来衡量S 符合自然语言的语法和语义规则的置信度。用一句简单的话说,统计语言模型就是计算一个句子的概率大小的这种模型。
n-gram模型可以减轻单词序列没有在训练集中出现过而引起的问题,即数据稀疏问题
n-gram模型问题 对于n-gram模型的问题,这两页ppt说的很明白
N-gram模型基于这样一种假设,当前词的出现只与前面N-1个词相关,而与其它任何词都不相关,整句的概率就是各个词出现概率的乘积。这些概率可以通过直接从语料中统计N个词同时出现的次数得到。常用的是二元的Bi-Gram(N=2)和三元的Tri-Gram(N=3).Bi-Gram所满足的假设是马尔科夫假设。
一般常用的N-Gram模型是Bi-Gram和Tri-Gram。分别用公式表示如下:
Bi-Gram:  P(T)=p(w1|begin) p(w2|w1) p(w3|w2)***p(wn|wn-1)
Tri-Gram:  P(T)=p(w1|begin1,begin2) p(w2|w1,begin1) p(w3|w2w1)***p(wn|wn-1,wn-2)
注意上面概率的计算方法:P(w1|begin)=以w1为开头的所有句子/句子总数;p(w2|w1)=w1,w2同时出现的次数/w1出现的次数。以此类推。
对于其中每项的计算举个例子:
由上可见Bi-Gram计算公式中的begin一般都是加个标签。
N-gram存在的问题:
举一个小数量的例子进行辅助说明:假设我们有一个语料库(注意语料库),如下:
老鼠真讨厌,老鼠真丑,你爱老婆,我讨厌老鼠。
想要预测“我爱老”这一句话的下一个字。我们分别通过 bigram 和 trigram 进行预测。
1)通过 bigram,便是要对 P(w|老)进行计算,经统计,“老鼠”出现了3次,“老婆”出现了1次,通过最大似然估计可以求得P(鼠|老)=0.75,P(婆|老)=0.25, 因此我们通过 bigram 预测出的整句话为: 我爱老鼠。
2)通过 trigram,便是要对便是要对 P(w|爱老)进行计算,经统计,仅“爱老婆”出现了1次,通过最大似然估计可以求得 P(婆|爱 老)=1,因此我们通过trigram 预测出的整句话为: 我爱老婆。显然这种方式预测出的结果更加合理。
问题一:随着 n 的提升,我们拥有了更多的前置信息量,可以更加准确地预测下一个词。但这也带来了一个问题,当N过大时很容易出现这样的状况:某些n-gram从未出现过, 导致很多预测概率结果为0, 这就是稀疏问题。 实际使用中往往仅使用 bigram 或 trigram 。(这个问题可以通过平滑来缓解参考: https://mp.weixin.qq.com/s/NvwB9H71JUivFyL_Or_ENA )
问题二:同时由于上个稀疏问题还导致N-gram无法获得上下文的长时依赖。
问题三:n-gram 基于频次进行统计,没有足够的泛化能力。

n-gram总结:统计语言模型就是计算一个句子的概率值大小,整句的概率就是各个词出现概率的乘积,概率值越大表明该句子越合理。N-gram是典型的统计语言模型,它做出了一种假设,当前词的出现只与前面N-1个词相关,而与其它任何词都不相关,整句的概率就是各个词出现概率的乘积。它其中存在很多问题,再求每一个词出现的概率时,随着N的提升,能够拥有更多的前置信息量,可以使得当前词的预测更加准确,但是当N过大时会出现稀疏问题,导致很多词的概率值为0,为解决这一问题,因此常用的为bigram 或 trigram,这就导致N-gram无法获得上文的长时依赖。另一方面N-gram 只是基于频次进行统计,没有足够的泛化能力。
神经网络语言模型
2003年 Bengio 提出,神经网络语言模型( neural network language model, NNLM)的思想是提出词向量的概念,代替 ngram 使用离散变量(高维),采用连续变量(具有一定维度的实数向量)来进行单词的分布式表示,解决了维度爆炸的问题,同时通过词向量可获取词之间的相似性。
结合下图可知它所建立的语言模型的任务是根据窗口大小内的上文来预测下一个词,因此从另一个角度看它就是一个使用神经网络编码的n-gram模型。
它是一个最简单的神经网络,仅由四层构成,输入层、嵌入层、隐藏层、输出层。(从另一个角度看它就是一个使用神经网络编码的n-gram模型)
输入是单词序列的index序列,例如单词‘这’在字典(大小为∣V∣)中的index是10,单词‘是’的 index 是23,‘测’的 index 是65,则句子“这是测试”通过‘这是测’预测‘试’,窗口大小内上文词的index序列就是 10, 23, 65。嵌入层(Embedding)是一个大小为∣V∣×K的矩阵(注意:K的大小是自己设定的,这个矩阵相当于随机初始化的词向量,会在bp中进行更新,神经网络训练完成之后这一部分就是词向量),从中取出第10、23、65行向量拼成3×K的矩阵就是Embedding层的输出了。隐层接受拼接后的Embedding层输出作为输入,以tanh为激活函数,最后送入带softmax的输出层,输出概率,优化的目标是使得待预测词其所对应的softmax值最大。
缺点:因为这是通过前馈神经网络来训练语言模型,缺点显而易见就是其中的参数过多计算量较大,同时softmax那部分计算量也过大。另一方面NNLM直观上看就是使用神经网络编码的 n-gram 模型,也无法解决长期依赖的问题。
RNNLM
它是通过RNN及其变种网络来训练语言模型,任务是通过上文来预测下一个词,它相比于NNLM的优势在于所使用的为RNN,RNN在处理序列数据方面具有天然优势, RNN 网络打破了上下文窗口的限制,使用隐藏层的状态概括历史全部语境信息,对比 NNLM 可以捕获更长的依赖,在实验中取得了更好的效果。RNNLM 超参数少,通用性更强;但由于 RNN 存在梯度弥散问题,使得其很难捕获更长距离的依赖信息。
Word2vec中的CBOW 以及skip-gram,其中CBOW是通过窗口大小内的上下文预测中心词,而skip-gram恰恰相反,是通过输入的中心词预测窗口大小内的上下文。
Glove 是属于统计语言模型,通过统计学知识来训练词向量
ELMO 通过使用多层双向的LSTM(一般都是使用两层)来训练语言模型,任务是利用上下文来预测当前词,上文信息通过正向的LSTM获得,下文信息通过反向的LSTM获得,这种双向是一种弱双向性,因此获得的不是真正的上下文信息。
GPT是通过Transformer来训练语言模型,它所训练的语言模型是单向的,通过上文来预测下一个单词
BERT通过Transformer来训练MLM这种真正意义上的双向的语言模型,它所训练的语言模型是根据上下文来预测当前词。
以上部分的详细介绍在NLP之预训练篇中有讲到
语言模型的评判指标
具体参考: https://blog.csdn.net/index20001/article/details/78884646
Perplexity可以认为是average branch factor(平均分支系数),即预测下一个词时可以有多少种选择。别人在作报告时说模型的PPL下降到90,可以直观地理解为,在模型生成一句话时下一个词有90个合理选择,可选词数越少,我们大致认为模型越准确。这样也能解释,为什么PPL越小,模型越好。 一般用困惑度Perplexity(PPL)衡量语言模型的好坏,困惑度越小则模型生成一句话时下一个词的可选择性越少,句子越确定则语言模型越好。
简单介绍 Word2vec是一种有效创建词嵌入的方法,它自2013年以来就一直存在。但除了作为词嵌入的方法之外,它的一些概念已经被证明可以有效地创建推荐引擎和理解时序数据。在商业的、非语言的任务中。
背景 由于任何两个不同词的one-hot向量的余弦相似度都为0,多个不同词之间的相似度难以通过onehot向量准确地体现出来。 word2vec?具的提出正是为了解决上?这个问题。它将每个词表?成?个定?的向量,并使得这些向量能较好地表达不同词之间的相似和类?关系。
word2vec模型 word2vec?具包含了两个模型,即跳字模型(skip-gram)和连续词袋模型(continuous bag of words,CBOW)。word2vec的input/output都是将单词作为one-hot向量来表示,我们可以把word2vec认为是词的无监督学习的降维过程。
MaxEnt 模型(最大熵模型): 可以使用任意的复杂相关特征,在性能上最大熵分类器超过了 Byaes 分类器。但是,作为一种分类器模型,这两种方法有一个共同的缺点:每个词都是单独进行分类的,标记(隐状态)之间的关系无法得到充分利用,具有马尔可夫链的 HMM 模型可以建立标记之间的马尔可夫关联性,这是最大熵模型所没有的。
最大熵模型的优点:首先,最大熵统计模型获得的是所有满足约束条件的模型中信息熵极大的模型;其次,最大熵统计模型可以灵活地设置约束条件,通过约束条件的多少可以调节模型对未知数据的适应度和对已知数据的拟合程度;再次,它还能自然地解决统计模型中参数平滑的问题。
最大熵模型的不足:首先,最大熵统计模型中二值化特征只是记录特征的出现是否,而文本分类需要知道特征的强度,因此,它在分类方法中不是最优的;其次,由于算法收敛的速度较慢,所以导致最大熵统计模型它的计算代价较大,时空开销大;再次,数据稀疏问题比较严重。
CRF(conditional random field) 模型(条件随机场模型):首先,CRF 在给定了观察序列的情况下,对整个的序列的联合概率有一个统一的指数模型。一个比较吸引人的特性是其为一个凸优化问题。其次,条件随机场模型相比改进的隐马尔可夫模型可以更好更多的利用待识别文本中所提供的上下文信息以得更好的实验结果。并且有测试结果表明:在采用相同特征集合的条件下,条件随机域模型较其他概率模型有更好的性能表现。
CRF 可以用于构造在给定一组输入随机变量的条件下,另一组输出随机变量的条件概率分布模型。经常被用于序列标注,其中包括词性标注,分词,命名实体识别等领域。
建一个条件随机场,我们首先要定义一个特征函数集,每个特征函数都以整个句子s,当前位置i,位置i和i-1的标签为输入。然后为每一个特征函数赋予一个权重,然后针对每一个标注序列l,对所有的特征函数加权求和,必要的话,可以把求和的值转化为一个概率值。
CRF 具有很强的推理能力,并且能够使用复杂、有重叠性和非独立的特征进行训练和推理,能够充分地利用上下文信息作为特征,还可以任意地添加其他外部特征,使得模型能够 获取的信息非常丰富。
CRF 模型的不足:首先,通过对基于 CRF 的结合多种特征的方法识别英语命名实体的分析,发现在使用 CRF 方法的过程中,特征的选择和优化是影响结果的关键因素,特征选择问题的好与坏,直接决定了系统性能的高低。其次,训练模型的时间比 MaxEnt 更长,且获得的模型很大,在一般的 PC 机上无法运行。
潜在语义分析(Latent Semantic Analysis,LSA)模型 在潜在语义分析(LSA)模型首先给出了这样一个 ‘‘分布式假设” :一个 单词的属性是由它所处的环境刻画的。这也就意味着如果两个单词在含义上比较接近,那么它们也会出现在相似的文本中,也就是说具有相似的上下文。 LSA模型在构建好了单词-文档矩阵之后,出于以下几种可能的原因,我们会使用奇异值分解(Singular Value Decomposition,SVD) 的方法来寻找该矩阵的一个低阶近似。
概率潜在语义分析(Probability Latent Semantic Analysis ,PLSA)模型 概率潜在语义分析(PLSA)模型其实是为了克服潜在语义分析(LSA)模型存在的一些缺点而被提出的。LSA 的一个根本问题在于,尽管我们可以把 U k 和 V k 的每一列都看成是一个话题,但是由于每一列的值都可以看成是几乎没有限制的实数值,因此我们无法去进一步解释这些值到底是什么意思,也更无法从概率的角度来理解这个模型。 PLSA模型则通过一个生成模型来为LSA赋予了概率意义上的解释。该模型假设,每一篇文档都包含一系列可能的潜在话题,文档中的每一个单词都不是凭空产生的,而是在这些潜在的话题的指引下通过一定的概率生成的。
在 PLSA 模型里面,话题其实是一种单词上的概率分布,每一个话题都代表着一个不同的单词上的概率分布,而每个文档又可以看成是话题上的概率分布。每篇文档就是通过这样一个两层的概率分布生成的,这也正是PLSA 提出的生成模型的核心思想。
PLSA 通过下面这个式子对d和 w 的联合分布进行了建模:

该模型中的 *z * 的数量是需要事先给定的一个超参数。需要注意的是,上面这 个式子里面给出了 P (w, d ) 的两种表达方式,在前一个式子里, *d * 和 w 都是在给定 *z * 的前提下通过条件概率生成出来的,它们的生成方式是相似的,因此是 ‘‘对称’’ 的;在后一个式子里,首先给定 d ,然后根据 P ( z | d ) 生成可能的话题 z ,然后再根据 P (w| z ) 生成可能的单词 w,由于在这个式子里面单词和文档的生成并不相似, 所以是 ‘‘非对称’’ 的。

上图给出了 PLSA 模型中非对称形式的 Plate Notation表示法。其中d表示 一篇文档,z 表示由文档生成的一个话题,w 表示由话题生成的一个单词。 在这个模型中, d和w 是已经观测到的变量,而z是未知的变量(代表潜在的话题)。
容易发现,对于一个新的文档而言,我们无法得知它对应的 P ( d ) 究竟是什么, 因此尽管 PLSA 模型在给定的文档上是一个生成模型,它却无法生成新的未知的文档。该模型的另外的一个问题在于,随着文档数量的增加, P ( z | d ) 的参数也会随着线性增加,这就导致无论有多少训练数据,都容易导致模型的过拟合问题。这两点成为了限制 PLSA 模型被更加广泛使用的两大缺陷。
潜在狄利克雷分配(Latent Dirichlet Analysis , LDA)模型
为了解决 PLSA 模型中出现的过拟合问题,潜在狄利克雷分配(LDA)模型被 Blei 等人提出,这个模型也成为了主题模型这个研究领域内应用最为广泛的模 型。LDA就是在PLSA的基础上加层贝叶斯框架,即LDA就是PLSA的贝叶斯版本(正因为LDA被贝叶斯化了,所以才需要考虑历史先验知识,才加的两个先验参数)。
从上一节我们可以看到,在 PLSA 这个模型里,对于一个未知的新文档 d ,我们对于 P ( d ) 一无所知,而这个其实是不符合人的经验的。或者说,它没有去使用本来可以用到的信息,而这部分信息就是 LDA 中所谓的先验信息。
具体来说,在 LDA 中,首先每一个文档都被看成跟有限个给定话题中的每一个存在着或多或少的关联性,而这种关联性则是用话题上的概率分布来刻画的, 这一点与 PLSA 其实是一致的。
但是在 LDA 模型中,每个文档关于话题的概率分布都被赋予了一个先验分布,这个先验一般是用稀疏形式的狄利克雷分布表示的。 这种稀疏形式的狄利克雷先验可以看成是编码了人类的这样一种先验知识:一般而言,一篇文章的主题更有可能是集中于少数几个话题上,而很少说在单独一篇文章内同时在很多话题上都有所涉猎并且没有明显的重点。
此外,LDA 模型还对一个话题在所有单词上的概率分布也赋予了一个稀疏形式的狄利克雷先验,它的直观解释也是类似的:在一个单独的话题中,多数情况是少部分(跟这个话题高度相关的)词出现的频率会很高,而其他的词出现的频率则明显较低。这样两种先验使得 LDA 模型能够比 PLSA 更好地刻画文档-话题-单词这三者的关系。
事实上,从 PLSA 的结果上来看,它实际上相当于把 LDA 模型中的先验分布转变为均匀分布,然后对所要求的参数求最大后验估计(在先验是均匀分布的前提下,这也等价于求参数的最大似然估计) ,而这也正反映出了一个较为合理的先验对于建模是非常重要的。
分词就是将连续的字序列按照一定的规范重新组合成词序列的过程。 现有的分词算法可分为三大类:基于字符串匹配的分词方法、基于理解的分词方法和基于统计的分词方法。 按照是否与词性标注过程相结合,又可以分为单纯分词方法和分词与标注相结合的一体化方法。
中文分词根据实现原理和特点,主要分为以下2个类别:
(1)基于词典分词算法 也称字符串匹配分词算法。该算法是按照一定的策略将待匹配的字符串和一个已建立好的“充分大的”词典中的词进行匹配,若找到某个词条,则说明匹配成功,识别了该词。常见的基于词典的分词算法分为以下几种:正向最大匹配法、逆向最大匹配法和双向匹配分词法等。 基于词典的分词算法是应用最广泛、分词速度最快的。很长一段时间内研究者都在对基于字符串匹配方法进行优化,比如最大长度设定、字符串存储和查找方式以及对于词表的组织结构,比如采用TRIE索引树、哈希索引等。
(2)基于统计的机器学习算法 这类目前常用的是算法是HMM、CRF(条件随机场)、SVM、深度学习等算法,比如stanford、Hanlp分词工具是基于CRF算法。以CRF为例,基本思路是对汉字进行标注训练,不仅考虑了词语出现的频率,还考虑上下文,具备较好的学习能力,因此其对歧义词和未登录词的识别都具有良好的效果。
常见的分词器都是使用机器学习算法和词典相结合,一方面能够提高分词准确率,另一方面能够改善领域适应性。
随着深度学习的兴起,也出现了 基于神经网络的分词器 ,例如有人员尝试使用双向LSTM+CRF实现分词器, 其本质上是序列标注 ,所以有通用性,命名实体识别等都可以使用该模型,据报道其分词器字符准确率可高达97.5%。算法框架的思路与论文《Neural Architectures for Named Entity Recognition》类似,利用该框架可以实现中文分词,如下图所示:
首先对语料进行字符嵌入,将得到的特征输入给双向LSTM,然后加一个CRF就得到标注结果。
目前中文分词难点主要有三个: 1、分词标准 :比如人名,在哈工大的标准中姓和名是分开的,但在Hanlp中是合在一起的。这需要根据不同的需求制定不同的分词标准。
2、歧义 :对同一个待切分字符串存在多个分词结果。 歧义又分为组合型歧义、交集型歧义和真歧义三种类型。
一般在搜索引擎中,构建索引时和查询时会使用不同的分词算法。常用的方案是,在索引的时候使用细粒度的分词以保证召回,在查询的时候使用粗粒度的分词以保证精度。
3、新词 :也称未被词典收录的词,该问题的解决依赖于人们对分词技术和汉语语言结构的进一步认识。
典型的文本分类过程可以分为三个步骤: 1. 文本表示(Text Representation) 这一过程的目的是把文本表示成分类器能够处理的形式。最常用的方法是向量空间模型,即把文本集表示成词-文档矩阵,矩阵中每个元素代表了一个词在相应文档中的权重。选取哪些词来代表一个文本,这个过程称为特征选择。常见的特征选择方法有文档频率、信息增益、互信息、期望交叉熵等等。为了降低分类过程中的计算量,常常还需要进行降维处理,比如LSI。 2. 分类器构建(Classifier Construction) 这一步骤的目的是选择或设计构建分类器的方法。不同的方法有各自的优缺点和适用条件,要根据问题的特点来选择一个分类器。我们会在后面专门讲述常用的方法。选定方法之后,在训练集上为每个类别构建分类器,然后把分类器应用于测试集上,得到分类结果。 3. 效果评估(Classifier Evaluation) 在分类过程完成之后,需要对分类效果进行评估。评估过程应用于测试集(而不是训练集)上的文本分类结果,常用的评估标准由IR领域继承而来,包括查全率、查准率、F1值等等。
1. Rocchio方法 每一类确定一个中心点(centroid),计算待分类的文档与各类代表元间的距离,并作为判定是否属于该类的判据。Rocchio方法的特点是容易实现,效率高。缺点是受文本集分布的影响,比如计算出的中心点可能落在相应的类别之外。
2. 朴素贝叶斯(na?ve bayes)方法 将概率论模型应用于文档自动分类,是一种简单有效的分类方法。使用贝叶斯公式,通过先验概率和类别的条件概率来估计文档对某一类别的后验概率,以此实现对此文档所属类别的判断。
3. K近邻(K-Nearest Neightbers, KNN)方法 从训练集中找出与待分类文档最近的k个邻居(文档),根据这k个邻居的类别来决定待分类文档的类别。KNN方法的优点是不需要特征选取和训练,很容易处理类别数目多的情况,缺点之一是空间复杂度高。KNN方法得到的分类器是非线性分类器。
4. 支持向量机(SVM)方法 对于某个类别,找出一个分类面,使得这个类别的正例和反例落在这个分类面的两侧,而且这个分类面满足:到最近的正例和反例的距离相等,而且是所有分类面中与正例(或反例)距离最大的一个分类面。SVM方法的优点是使用很少的训练集,计算量小;缺点是太依赖于分类面附近的正例和反例的位置,具有较大的偏执。
文本聚类过程可以分为3个步骤: 1. 文本表示(Text Representation) 把文档表示成聚类算法可以处理的形式。所采用的技术请参见文本分类部分。 2. 聚类算法选择或设计(Clustering Algorithms) 算法的选择,往往伴随着相似度计算方法的选择。在文本挖掘中,最常用的相似度计算方法是余弦相似度。聚类算法有很多种,但是没有一个通用的算法可以解决所有的聚类问题。因此,需要认真研究要解决的问题的特点,以选择合适的算法。后面会有对各种文本聚类算法的介绍。 3. 聚类评估(Clustering Evaluation) 选择人工已经分好类或者做好标记的文档集合作为测试集合,聚类结束后,将聚类结果与已有的人工分类结果进行比较。常用评测指标也是查全率、查准率及F1值。
1.层次聚类方法 层次聚类可以分为两种:凝聚(agglomerative)层次聚类和划分(divisive)层次聚类。凝聚方法把每个文本作为一个初始簇,经过不断的合并过程,最后成为一个簇。划分方法的过程正好与之相反。层次聚类可以得到层次化的聚类结果,但是计算复杂度比较高,不能处理大量的文档。
2.划分方法 k-means算法是最常见的划分方法。给定簇的个数k,选定k个文本分别作为k个初始簇,将其他的文本加入最近的簇中,并更新簇的中心点,然后再根据新的中心点对文本重新划分;当簇不再变化时或经过一定次数的迭代之后,算法停止。k-means算法复杂度低,而且容易实现,但是对例外和噪声文本比较敏感。另外一个问题是,没有一个好的办法确定k的取值。
3.基于密度的方法 为了发现任意形状的聚类结果,提出了基于密度的方法。这类方法将簇看作是数据空间中被低密度区域分割开的高密度区域。常见的基于密度的方法有DBSCAN, OPTICS, DENCLUE等等。
4.神经网络方法 神经网络方法将每个簇描述为一个标本,标本作为聚类的"原型",不一定对应一个特定的数据,根据某些距离度量,新的对象被分配到与其最相似的簇中。比较著名的神经网络聚类算法有:竞争学习(competitive learing)和自组织特征映射(self-organizing map)[Kohonen, 1990]。神经网络的聚类方法需要较长的处理时间和复杂的数据复杂性,所以不适用于大型数据的聚类。

帮我弄一个英文的拿破仑简介呗

Napoleon Bonaparte was born on August 15, 1769 in Ajaccio on the Mediterranean island of Corsica, the son of Carlo and Letizia Bonaparte. Through his military exploits and his ruthless efficiency, Napoleon rose from obscurity to become Napoleon I, Empereur des Francais (Emperor of the French). He is both a historical figure and a legend—and it is sometimes difficult to separate the two. The events of his life fired the imaginations of great writers, film makers, and playwrights whose works have done much to create the Napoleonic legend.
]Napoleon decided on a military career when he was a child, winning a scholarship to a French military academy at age 14. His meteoric rise shocked not only France but all of Europe, and his military conquests threatened the stability of the world.
Napoleon was one of the greatest military commanders in history. He has also been portrayed as a power hungry conqueror. Napoleon denied those accusations. He argued that he was building a federation of free peoples in a Europe united under a liberal government. But if this was his goal, he intended to achieve it by taking power in his own hands. However, in the states he created, Napoleon granted constitutions, introduced law codes, abolished feudalism, created efficient governments and fostered education, science, literature and the arts.
Emperor Napoleon proved to be an excellent civil administrator. One of his greatest achievements was his supervision of the revision and collection of French law into codes. The new law codes—seven in number—incorporated some of the freedoms gained by the people of France during the French revolution, including religious toleration and the abolition of serfdom. The most famous of the codes, the Code Napoleon or Code Civil, still forms the basis of French civil law. Napoleon also centralized France's government by appointing prefects to administer regions called departments, into which France was divided.
While Napoleon believed in government "for" the people, he rejected government "by" the people. His France was a police state with a vast network of secret police and spies. The police shut down plays containing any hint of disagreement or criticism of the government. The press was controlled by the state. It was impossible to express an opinion without Napoleon's approval.
Napoleon I ——Emperor of the French
1769 - 1821
“Ambition is never content, even on the summit of greatness.”
——Napoleon Bonaparte
--------------------------------------------------------------------------------
Napoleon Bonaparte was born on August 15, 1769 in Ajaccio on the Mediterranean island of Corsica. Through his military exploits and his ruthless efficiency, Napoleon rose from obscurity to become Napoleon I, Empereur des Francais (Emperor of the French). He is both a historical figure and a legend—and it is sometimes difficult to separate the two. The events of his life fired the imaginations of great writers, film makers, and playwrights whose works have done much to create the Napoleonic legend.
Napoleon decided on a military career when he was a child, winning a scholarship to a French military academy. His meteoric rise shocked not only France but all of Europe, and his military conquests threatened the stability of the world.
Napoleon was one of the greatest military commanders in history. He has also been portrayed as a power hungry conqueror. Napoleon denied being such a conqueror. He argued that he was building a federation of free peoples in a Europe united under a liberal government. But if this was his goal, he intended to achieve it by taking power in his own hands. However, in the states he created, Napoleon granted constitutions, introduced law codes, abolished feudalism, created efficient governments and fostered education, science, literature and the arts.
Emperor Napoleon proved to be an excellent civil administrator. One of his greatest achievements was his supervision of the revision and collection of French law into codes. The new law codes—seven in number—incorporated some of the freedoms gained by the people of France during the French revolution, including religious toleration and the abolition of serfdom. The most famous of the codes, the Code Napoleon or Code Civil, still forms the basis of French civil law. Napoleon also centralized France's government by appointing prefects to administer regions called departments, into which France was divided.
While Napoleon believed in government "for" the people, he rejected government "by" the people. His France was a police state with a vast network of secret police and spies. The police shut down plays containing any hint of disagreement or criticism of the government. The press was controlled by the state. It was impossible to express an opinion without Napoleon's approval.
Napoleon's own opinion of his career is best stated in the following quotation:
“I closed the gulf of anarchy and brought order out of chaos. I rewarded merit regardless of birth or wealth, wherever I found it. I abolished feudalism and restored equality to all regardless of religion and before the law. I fought the decrepit monarchies of the Old Regime because the alternative was the destruction of all this. I purified the Revolution.”
His death on 5 May 1821 brought relief to the royal houses of Europe and it was only in 1840 that his body was allowed to return to his beloved France.
我的比较好,又简单,短。要我的如何?
August 15, 1769 birth.
1784 into the Royal Military Academy in Paris learning. The following year, Napoleon graduated to the title of second lieutenant.
1794 because of chaos-meritorious was appointed Major General, artillery brigade commander.
In 1796, only 26-year-old Napoleon was appointed military commander of France and Italy. A dilapidated military expeditionary Italy. Exploits prominent.
In 1799 Napoleon launched Brumaire coup, which overthrew the Governor government, the establishment of the three ruling.
On December 2, 1804, Napoleon at Notre Dame in Paris coronation ceremony, said Napoleon them forever.
Ostritz Jena's victory over the bonaparte become almost the entire European surnames
In 1812 invasion of Russia failed, is a turning point in its political life.
1814 abdication.
On March 20, 1815, Napoleon return to Paris, the establishment of a "hundred days."
In June 1815, French troops in the battle of Waterloo in the swamped, Napoleon second abdication.
On August 15, 1769 was born.
In 1784 changed over to the Paris king family military academy tostudy. Next year, Napoleon to second lieutenant will hold thegraduation.
In 1794 because put down a rebellion to have the merit to appoint asmajor general, artillery brigade commander.
In 1796, year only 26 year-old Napoleon is appointed as French Italyarmed force commanders. Declines army expedition Italy. Themeritorious military service is prominent.
In 1799 Napoleon started the fog month coup d'etat, overthrewsupervises the government, established three people to be in power.
On December 2, 1804, Napoleon held the coronation ceremony in theNotre Dame cathedral, called Napoleon th.
Oersted Ritz, the Ye Na big victory causes bonaparte to become thenearly entire European the surname
In 1812 invaded Russia to be defeated, was its politicalsurvival turning point.
In 1814 abdicated.
On March 20, 1815, Napoleon returned to Paris, the establishment "thehundred days dynasty".
In June, 1815, the law armed force in the battle of waterloo 覆没,Napoleon second time abdicated.
微软百科非常详尽的介绍:
Napoleon I
I INTRODUCTION
Napoleon I (1769-1821), emperor of the French, who consolidated and institutionalized many reforms of the French Revolution. One of the greatest military commanders of all time, he conquered the larger part of Europe and did much to modernize the nations he ruled.
Napoleon was born on August 15, 1769, in Ajaccio, Corsica, and was given the name Napoleone (in French his name became Napoleon Bonaparte). He was the second of eight children of Carlo (Charles) Buonaparte and Letizia Ramolino Buonaparte, both of the Corsican-Italian gentry. No Buonaparte had ever been a professional soldier. Carlo was a lawyer who had fought for Corsican independence, but after the French occupied the island in 1768, he served as a prosecutor and judge and entered the French aristocracy as a count. Through his father’s influence, Napoleon was educated at the expense of King Louis XVI, at Brienne and the école Militaire, in Paris. Napoleon graduated in 1785, at the age of 16, and joined the artillery as a second lieutenant.
After the Revolution began, he became a lieutenant colonel (1791) in the Corsican National Guard. In 1793, however, Corsica declared independence, and Bonaparte, a French patriot and a Republican, fled to France with his family. He was assigned, as a captain, to an army besieging Toulon, a naval base that, aided by a British fleet, was in revolt against the republic. Replacing a wounded artillery general, he seized ground where his guns could drive the British fleet from the harbor, and Toulon fell. As a result Bonaparte was promoted to brigadier general at the age of 24. In 1795 he saved the revolutionary government by dispersing an insurgent mob in Paris. In 1796 he married Joséphine de Beauharnais, the widow of an aristocrat guillotined in the Revolution and the mother of two children.
II EARLY CAMPAIGNS
Also in 1796, Bonaparte was made commander of the French army in Italy. He defeated four Austrian generals in succession, each with superior numbers, and forced Austria and its allies to make peace. The Treaty of Campo Formio provided that France keep most of its conquests. In northern Italy he founded the Cisalpine (Italian) Republic (later known as the kingdom of Italy) and strengthened his position in France by sending millions of francs worth of treasure to the government. In 1798, to strike at British trade with the East, he led an expedition to Ottoman-ruled Egypt, which he conquered. His fleet, however, was destroyed by the British admiral Horatio Nelson, leaving him stranded. Undaunted, he reformed the Egyptian government and law, abolishing serfdom and feudalism and guaranteeing basic rights. The French scholars he had brought with him began the scientific study of ancient Egyptian history. In 1799 he failed to capture Syria, but he won a smashing victory over the Ottomans at Abū Qīr (Abukir). France, meanwhile, faced a new coalition; Austria, Russia, and lesser powers had allied with Britain.
III NAPOLEONIC RULE IN FRANCE
Bonaparte, no modest soul, decided to leave his army and return to save France. In Paris, he joined a conspiracy against the government. In the coup d’etat of November 9-10, 1799 (18-19 Brumaire), he and his colleagues seized power and established a new regime—the Consulate. Under its constitution, Bonaparte, as first consul, had almost dictatorial powers. The constitution was revised in 1802 to make Bonaparte consul for life and in 1804 to create him emperor. Each change received the overwhelming assent of the electorate. In 1800, he assured his power by crossing the Alps and defeating the Austrians at Marengo. He then negotiated a general European peace that established the Rhine River as the eastern border of France. He also concluded an agreement with the pope (the Concordat of 1801), which contributed to French domestic tranquillity by ending the quarrel with the Roman Catholic church that had arisen during the Revolution. In France the administration was reorganized, the court system was simplified, and all schools were put under centralized control. French law was standardized in the Code Napoléon, or civil code, and six other codes. They guaranteed the rights and liberties won in the Revolution, including equality before the law and freedom of religion.
IV WARS OF CONQUEST
In April 1803 Britain, provoked by Napoleon’s aggressive behavior, resumed war with France on the seas; two years later Russia and Austria joined the British in a new coalition. Napoleon then abandoned plans to invade England and turned his armies against the Austro-Russian forces, defeating them at the Battle of Austerlitz on December 2, 1805. In 1806 he seized the kingdom of Naples and made his elder brother Joseph king, converted the Dutch Republic into the kingdom of Holland for his brother Louis, and established the Confederation of the Rhine (most of the German states) of which he was protector. Prussia then allied itself with Russia and attacked the confederation. Napoleon destroyed the Prussian army at Jena and Auerst?dt (1806) and the Russian army at Friedland. At Tilsit (July 1807), Napoleon made an ally of Tsar Alexander I and greatly reduced the size of Prussia (see Tilsit, Treaty of). He also added new states to the empire: the kingdom of Westphalia, under his brother Jerome, the duchy of Warsaw, and others.
Napoleon had meanwhile established the Continental System, a French-imposed blockade of Europe against British goods, designed to bankrupt what he called the “nation of shopkeepers.” In 1807 Napoleon seized Portugal. In 1808, he made his brother Joseph king of Spain, awarding Naples to his brother-in-law, Joachim Murat. Joseph’s arrival in Spain touched off a rebellion there, which became known as the Peninsular War. Napoleon appeared briefly and scored victories, but after his departure the fighting continued for five years, with the British backing Spanish armies and guerrillas. The Peninsular War cost France 300,000 casualties and untold sums of money and contributed to the eventual weakening of the Napoleonic empire.
In 1809 Napoleon beat the Austrians again at Wagram, annexed the Illyrian Provinces (now part of Slovenia, Croatia, Bosnia and Herzegovina, and the Federal Republic of Yugoslavia), and abolished the Papal States. He also divorced Joséphine, and in 1810 he married the Habsburg archduchess Marie Louise, daughter of the Austrian emperor. By thus linking his dynasty with the oldest ruling house in Europe, he hoped that his son, who was born in 1811, would be more readily accepted by established monarchs. In 1810 also, the empire reached its widest extension with the annexation of Bremen, Lübeck, and other parts of north Germany, together with the entire kingdom of Holland, following the forced abdication of Louis Bonaparte.
V NAPOLEONIC RULE IN EUROPE
In all the new kingdoms created by the emperor, the Code Napoléon was established as law. Feudalism and serfdom were abolished, and freedom of religion established (except in Spain). Each state was granted a constitution, providing for universal male suffrage and a parliament and containing a bill of rights. French-style administrative and judicial systems were required. Schools were put under centralized administration, and free public schools were envisioned. Higher education was opened to all who qualified, regardless of class or religion. Every state had an academy or institute for the promotion of the arts and sciences. Incomes were provided for eminent scholars, especially scientists. Constitutional government remained only a promise, but progress and increased efficiency were widely realized. Not until after Napoleon’s fall did the common people of Europe, alienated from his governments by war taxes and military conscription, fully appreciate the benefits he had given them.
VI NAPOLEON’S DOWNFALL
In 1812 Napoleon, whose alliance with Alexander I had disintegrated, launched an invasion of Russia that ended in a disastrous retreat from Moscow. Thereafter all Europe united against him, and although he fought on, and brilliantly, the odds were impossible. In April 1814, his marshals refused to continue the struggle. After the allies had rejected his stepping down in favor of his son, Napoleon abdicated unconditionally and was exiled to the Mediterranean island of Elba. Marie Louise and his son were put in the custody of her father, the emperor of Austria. Napoleon never saw either of them again. Napoleon himself, however, soon made a dramatic comeback. In March 1815, he escaped from Elba, reached France, and marched on Paris, winning over the troops sent to capture him. In Paris, he promulgated a new and more democratic constitution, and veterans of his old campaigns flocked to his support. Napoleon asked peace of the allies, but they outlawed him, and he decided to strike first. The result was a campaign into Belgium, which ended in defeat at the Battle of Waterloo on June 18, 1815. In Paris, crowds begged him to fight on, but the politicians withdrew their support. Napoleon fled to Rochefort, where he surrendered to the captain of the British battleship Bellerophon. He was then exiled to Saint Helena, a remote island in the south Atlantic Ocean, where he remained until his death on May 5, 1821.
VII THE NAPOLEONIC LEGEND
The cult of Napoleon as the “man of destiny” began during his lifetime. In fact, he had begun to cultivate it during his first Italian campaign by systematically publicizing his victories. As first consul and emperor, he had engaged the best writers and artists of France and Europe to glorify his deeds and had contributed to the cult himself by the elaborate ceremonies with which he celebrated his rule, picturing himself as the architect of France’s greatest glory. He maintained that he had preserved the achievements of the Revolution in France and offered their benefits to Europe. His goal, he said, was to found a European state—a “federation of free peoples.” Whatever the truth of this, he became the arch-hero of the French and a martyr to the world. In 1840 his remains were returned to Paris at the request of King Louis-Philippe and interred with great pomp and ceremony in the Invalides, where they still lie.
VIII EVALUATION
Napoleon’s influence is evident in France even today. Reminders of him dot Paris—the most obvious being the Arc de Triomphe, the centerpiece of the city, which was built to commemorate his victories. His spirit pervades the constitution of the Fifth Republic; the country’s basic law is still the Code Napoléon, and the administrative and judicial systems are essentially Napoleonic. A uniform state-regulated system of education persists. Napoleon’s radical reforms in all parts of Europe cultivated the ground for the revolutions of the 19th century. Today, the impact of the Code Napoléon is apparent in the law of all European countries.
Napoleon was a driven man, never secure, never satisfied. “Power is my mistress,” he said. His life was work-centered; even his social activities had a purpose. He could bear amusements or vacations only briefly. His tastes were for coarse food, bad wine, cheap snuff. He could be charming—hypnotically so—for a purpose. He had intense loyalties—to his family and old associates. Nothing and no one, however, were allowed to interfere with his work.
Napoleon was sometimes a tyrant and always an authoritarian, but one who believed in ruling by mandate of the people, expressed in plebiscites. He was also a great enlightened monarch—a civil executive of enormous capacity who changed French institutions and tried to reform the institutions of Europe and give the Continent a common law. Few deny that he was a military genius. At Saint Helena, he said, “Waterloo will erase the memory of all my victories.” He was wrong; for better or worse, he is best remembered as a general, not for his enlightened government, but the latter must be counted if he is justly to be called Napoleon the Great.
See French Revolution; Napoleonic Wars. See also separate articles on individual battles mentioned.
Contributed By:
Owen Connelly
Microsoft ? Encarta ? Encyclopedia 2003. ? 1993-2002 Microsoft Corporation. All rights reserved.

阅读更多 >>>  suggestion,suggestion什么意思

外研社小学英语四年级上册教案三篇

【 #教案# 导语】四年级的学生接触英语已有一年之久,对英语学习的积极性与参与性较高,好动好胜,喜爱小组比赛。同时他们在认读和自学方面比起三年级均有一定程度的提高,能够听音指读、听音选择、听音填空等,能够认读基本单词,并能借助图文了解课文大意。 准备了以下内容,供大家参考! 篇一 《seventeen.eighteen..》   课题累型:
  新授
  教学目标:
  1、基础目标
  (1)学生能听、说和认读thirteen、fourteen、fifteen、sixteen、seventeen、eighteen、nineteen、twenty八个数字单词,并做到发音正确、规范。
  (2)帮助学生在学习新知识的过程中复习旧知识,并能结合已学句型“How many……can you see ?/ Do you have?/ I can see……/ I have……/等句型进行回答、替换和运用。
  2、拓展目标
  (1)将单词教学融入句型及对话教学之中,鼓励学生在实际情景中恰当地运用已学句型。
  (2)培养学生的注意力和观察力,发展学生的发散思维,激发学生的学习兴趣,培养学生的协作精神。
  教学重点:
  掌握八个新授数字单词,并能结合已学知识在实际情景中运用。
  教学难点:
  thirteen、 fifteen、 eighteen、 twenty的发音。
  教学准备:
  录音机和磁带,教师自己准备的一幅彩图画,数字单词卡片,在讲台上放15本书,一个盒子,18支铅笔,一个篮球。
  学生分析:
  本单元授课对象是农村小学四年级学生。经过一年多的学习,对英语保持着浓厚的兴趣 。在上册对1—12的数字的英语学习后,他们期待后面更大数字的学习。所以这一单元,特别是数字词汇的学习,他们会学得主动而轻松。因为教师对学习的了解,本单元就该更关注后进生,给他们更多的机会表现自己,激发他们的学习积极性,让所有学生自然地同步前进,并为接下去的学习打好基础。
  教学过程:
  一、Warm-up/Revision(热身、复习)
  1、Everyday English(每日英语)
  2、Sing a song.“Ten Little Candle Dance”(师生表演唱)
  3、Introduce themselves to new teachers (自我介绍)
  Eg:My name is……。I’m a girl/ boy.
  I’m from……I like……etc.
  4、Play a guessing game(猜数字)
  T:which number is in my pocket?
  二、Presentation(新知识呈现)
  1、Learn to say new words.(学习新数字)
  (1)数学练习,学习thirteen(13)
  A、从5+5=10,5+6=11,5+7=12,5+8=13,引出数字thirteen,并练习朗读该单词。(跟读,开火车读,齐读)
  B、Make 13 (让学生用英语编数学算式?+?=13)
  Eg:What’s six and seven? Six and seven is 13.
  C、用英语讲解西方国家对13的看法,结合幸运数字讲解西方人很忌讳13,并举例说明。
  (2)实物呈现,学习fifteen(15)
  A、引导学生利用身边的东西(如书本、同学、书包等)来数一数和说一说。
  Eg:T:Look!what are they? Ss:Books.
  T:How many books can you see?
  Let’s count,please.
  Ss:One,two……fifteen. B、练习朗读该单词(跟读,打行读,小组读,齐读)
  C、用15造句子。
  Eg:This is 15. Seven and eight is 15.
  I like 15. My brother/ sister is 15.
  I have fifteen……
  (3)猜一猜,学习eighteen(18)
  A、T:I have something in my bag. Guess. What do I have?
  Ss:……
  T:Look!What are they?
  Ss:Pencils.
  T:How many pencils do I have? Guess. Please!
  Ss:……
  T:Open it and see. Count please. One……eighteen. Oh, eighteen pencils. That’s right. I have eighteen pencils.
  B、练习朗读该单词(跟读,开火车两人读,小组读,齐读)
  C、用eighteen说词组。
  Eg:eighteen books
  eighteen dogs……
  (4)看一看,听一听,数一数,学习twenty(20)
  A、T:(拿着球)Look, What’s that? Oh, It’s a ball. I like balls. And I can bounce the ball. Now, I bounce the ball, please count!教师拍球,学生数数,数至20,然后故意把球拍丢了。
  B、练习朗读该单词(跟读,指读,男女读,齐读)
  (5)看图添画,学习14、16、17、19。
  教师也用简笔画边画苹果,边提问学生:What can you see? How many apples can you see?(本环节的教学内容可以从数字12添到20。再出示这几个数字卡片让学生读一读。)
  2、Listen and repeat.(先给学生听一遍录音,让学生复述自己听到的,然后再翻书指着单词跟读)
  3、小结规律,让学生看着黑板上8个新数字,读一读,你发现了什么?让学生明白13、15、18、20的读音是没有规律的,而14、16、17、19是比较简单,只要在4、6、7、9后加[ti:n]。
  三、Practice and consolidation(操练和巩固)
  1、明辩是非读单词。(教师指着黑板上的单词,如果指的和读的是一致的学生就跟,反之保持沉默)
  2、听音举卡片,老师说一个数字,学生要边讲边举起该数字卡片,比一比哪一组动作最快。
  Eg:T:Thirteen, thirteen, Where is 13?
  Ss:Thirteen,thirteen, Here it is.
  3、同桌游戏(划拳),两人各出示手指表示的数字再两个相加,谁先把答案说出来,谁就可赢得对方的卡片,看谁的卡片赢得的多为胜者,可用以下句型出示手指。
  Eg:How many, how many fingers?
  Five and seven is twelve. How old are you? How old are you? I’m____
  四、Extension(拓展)
  Talk about the picture
  教师出示一幅课前画好的彩图(类似活动手册第16页,图的量还可加大,丰富一下,还加上风筝、白云、小鸟、花草等,把颜色、数字、动植物都画进去)。
  教师引导学生就图展开讨论,教师先示范引导:Look this is a picture. Is it beautiful? What can you see? What colour? How many? ……
  让学生先小组讨论时尽量运用所学句型,最后让几对学生到讲台前谈论图画,展示他们的讨论结果。教师可适时板书几个句子。
  五、Homework(自选)
  1、一:(1)模仿录音朗读3遍,并制作自己的英语录音专集。
  (2)Listen and connect the dots.(听音连点)
  2、二:Read, think and write.(看、想、填)
  five fifteen six twelve eighteen
  seventeen thirteen nine
  3、三:Make a survey.(填写调查表,请用上How many…do you have?I have…。You have…)
  板书设计:
  Unit 1 …seventeen,eighteen,nineteen,twenty!
  A & B Let’s learn Let’s play
  5+6=11 5+7=12 5+8=13 thirteen
  4→14 fourteen
  15 fifteen
  6→16 sixteen
  7→17 seventeen How many……?
  18 eighteen They are……。
  9→19 nineteen How nice!
  20 twenty The orange one is a bird.    篇二 《There are twelve months.. 》   一、教材分析:学习7-12月份
  二、教学目标
  1、知识目标:
  1)基本能够听懂,会说,会读词汇:July, August, September, October, November, December.
  2)学会用“There are…birthdays in …”以及“There is one birthday in …”的语言结构。
  3)复习词汇:flowers, festival, birds, snow, rain, warm, hot, cool, cold, sunny.
  2 技能目标:谈论月份以及关于各个月的情况。
  3情感目标:从小培养学生爱护环境的意识。通过学习月份的歌曲,培养学生的乐感和美感。
  三、教学重点:词汇:July, August, September, October, November, December.
  “There are…birthdays in …”以及“There is one birthday in …”的语言结构。
  教学难点:November的读音。
  四、教学准备:录音机、磁带,教学卡片、挂图、日历等。
  五、Teaching steps:
  Step1: Warmer
  1. Greetings
  T: Hello, boys and girls. Ss: Hello, Miss Yi.
  T: How are you today? Ss: I’m fine , thanks . And how are you?
  T: I’m great. Thank you so much. Look! How is the weather today?
  Ss: It’s sunny today.
  T: How do you feel now? Are you happy? Ss: Yes. We’re so happy.
  T: Let’s sing a song , ok? Ss: OK!
  T& Ss: Spring is warm, spring is warm.
  Summer is hot, summer is hot.
  Autumn is cool, autumn is cool.
  Winter is cold, winter is cold.
  2. 复习1-6月份的英文。
  Step2: Presentation
  1、Lead in
  T: How many months are there in a year?
  Ss: There are twelve months in the year.(引出课题)
  板书:Unit 2 There are twelve months in the year.
  T: Let’s pass the toy dog and say the words from January to December.
  Ss: January, February, March, April, May, June….接着老师带着学生一起说: July, August, September, October, November and December. (在游戏的过程中引入新单词)
  T: Today, we’re going to learn the months.
  2、教授—July
  T: Boys and girls, let’s play a game, ok? Ss: OK!
  T: I will speak these names of the months. And you can show me your fingers to show. Ss: Yes.
  T: May. Ss: (show five fingers) May.
  T: March. Ss: (show three fingers) March.
  T: July. 提示学生show seven fingers. (July)
  T: Do you like July?
  S1: Yes. I like July. We can go to the park in July.
  3、教授—August
  T: How old are you?
  Ss: I’m 10 years old.
  T: When is your birthday?
  Ss:八月。
  T: Oh , your birthday is in August. Whose birthday is also in August?
  S2: My birthday is in August, too.
  T: How many birthdays are there in August?
  Ss: There are….
  4、教授-- September
  T:Boys and girls, when is Teacher’s Day? Who can tell me?
  S3: 九月十日。
  T: You’re so clever. September,10 th . Do you like September?
  S4: Yes. I like September .
  T: What can you do?
  S5: We can go to school.
  T: Sounds great.
  5、教授—October
  T:You can guess so well. Do you know which month it is? “ It’s cool and it’s in Autumn. There are 31 days in it. And there is National Day ,too.”
  S6: October.
  T: Yes, you are right.
  6、教授—November
  T: I like November. What do you like?
  S7: I like June.
  S8: I like March.
  T: Do you like November?
  Ss: Yes, we like November.
  7、教授—December
  T: Oh ,I’m a weather reporter now. In Beijing , it’s cold and there is snow in December . (播报天气预报)Can you hear which month it is?
  Ss: 十二月。
  T: December . Whose birthday is in December?
  S9: My birthday is in December.
  T: How many birthdays are there in December? Let’s count.
  Ss: There are ….birthdays in December.
  T: I heard Ms Smart are counting birthdays now. Please listen to the tape.
  Step3: Practice
  1. Listen and do.
  T: Boys and girls , please listen and do.
  T:1) Whose birthday is in July? Please Stand up.
  2) Whose birthday is in August? Put your hands on your head, please!
  3) Whose birthday is in September? Touch your ears, please!
  4) Whose birthday is in October? Clap, please!
  5) Whose birthday is in November? Point to the teacher, please !
  6) Whose birthday is in December? Laugh, please!
  2. Group work
  完成课本活动3:先请几个学生朗读对话示例,然后把学生分成小组,让他们根据图中给的信息展开对话,发表自己对各个月份的看法。
  e.g. S1: Do you like January?
  S2: Yes, I like January .
  S1: Why?
  S2: There are lots of festivals.
  S3: Which month do you like?
  S4: I like October.
  S3: Why?
  S4: We can fly kites.
  S5: Which month is your favourite?
  S6: My favourite month is June.
  S5: Why?
  S6: It’s warm. I can ride a bike
  Homework:1熟记1-12月份的单词
  2、询问父母,然后在课堂上互相介绍自己父母的情况。 篇三 《There is one birthday in may》   内容分析:
  本课是新标准英语(三年级起始)第三册第十模块,它的语言功能主要是运用There be句型去描述与月份相关的事情。第一单元以故事情境——Lingling与MsSmart谈论玲玲朋友生日个数的对话呈现,主要是在巩固There be句型的同时学习1——6月份的表达方式。第二单元则是继续描述7——12月份生日个数的情况,并谈论全年不同月份当地的天气状况以及喜欢的运动。本模块预计四课时完成,本课时为第一课时。
  情况分析:
  (1)四年级的学生接触英语已有一年之久,对英语学习的积极性与参与性较高,好动好胜,喜爱小组比赛。同时他们在认读和自学方面比起三年级均有一定程度的提高,能够听音指读、听音选择、听音填空等,能够认读基本单词,并能借助图文了解课文大意。
  (2)在本学期第一模块学生已经学习了数词1—100,对于How many? There be 句型在听力方面也有感知与理解。同时,课文中的单词march, birthday, happy等词语学生比较熟悉,结合对话语境能够较快理解课文的背景。
  教学目标:
  (1)知识能力目标
  ①能够寻找单词中发音相同的字母组合,以旧带新,听懂、辨别并认读1—6月份的月名,January\ February\March\April\May\June。
  ②能够在倾听完整语篇情境中看图了解对话内容,并了解、运用There be 句型描述数量。如,January. There are three.
  (2)情感态度目标
  ①培养学生对所学英语中接触的外国文化知识的兴趣。
  ②让学生通过课文的学习感受生日聚会的快乐,并能在自己快乐的同时表达对父母的爱。
  教学重点:
  (1)单词:January\February\March\April\May\June
  (2)句型:“There be”功能句的简单运用
  教学难点:
  (1)单词:January和 February单词比较长,记忆比较难;而April\June则是发音难点。
  (2)There is和 There are 学生对单数复数概念模糊,容易搞错。
  教学过程:
  一、Warmer
  1、Play a Game: Count count count数数看
  T claps hands and asks Ss “How many are there?”
  Ss listen ,count and try to say the number or the sentence “There is/ are…”
  【设计意图:用老师拍手,学生数数的活动作为热身,既能复习数字,又能在听力方面不断倾听句型There be,为学生在听后运用句型There is / There are做铺垫。】
  2、Lead in the evaluation
  (1)Put up the cards of the months(出示1—6个月的月份卡,排成桥型),let Ss count the cards and learn to say “There is/ are…”;(板书主要句型于黑板四线格上,并用不同色粉笔突出is和are)
  Ss count and learn to say and make out “There is ”and “There are”
  (2)Talk about the evaluation——分组PK
  【设计意图:巧妙呈现本节课的课堂评价方式:把月份卡(六个月份名的缩写制作成椭圆形卡片)排成桥形贴在黑板上,把全班分成两个小组,要求各小组通过课堂上的表现力争尽快从此岸到达彼岸。同时,通过月份卡的呈现,引导学生数数,并学习运用句型There be准确描述。】
  二、Lead in
  1、Lead in
  T:We’re counting just now. Lingling is counting, too. She is counting her friends’ birthdays.
  Present the picture of Lingling’s counting by CAI
  【设计意图:由我们自己的数数活动过渡到玲玲的数生日的事情,直接导入课文,简单明了,切中本节课的对话背景。】
  2、Listen, read and find the answer
  Ask Ss a question about Lingling’s counting ——“How many birthdays are there in a year?”,and let Ss open their books, then plays the tape
  Ss read the book, listen to the whole lesson and try to find the answer
  【设计意图:通过整体语篇的听力输入,要求学生们速听速读课文,整体感知对话情境,进而思考回答问题。培养学生认真倾听,认读思考的良好自学习惯。】
  3、Present the Topic
  Show Picture Two to Ss and present the task of this module
  Read the title together: ——Module 10 The Months
  Unit 1 There is one birthday in May.
  S三、Presentation and Practice
  1、Listen and fill in the blanks
  Ask Ss the questions about Picture Two——“How many birthdays are there in January/ February/ March/ April/ May/ June?”
  Ss listen to each sentence and fill in the blanks
  《随堂练兵卡》一、听音填空
  一月January. There are .
  二月February. There are .
  三月March. There is .
  四月April. There are .
  五月May. There is .
  六月June. There are .
  【设计意图:抓住课文的重点段落,让学生有针对性地倾听,再通过随堂练兵一听音填空这个听力练习任务,引导学生再次边听边精读课文第二段,完成听力任务,使阅读更具目标性和实效性。】
  2、Learn the months
  Present the months of this exercise paper(《随堂练兵卡》)by CAI and ask Ss “ 哪个月份生日个数最少?”
  Ss read the paper(《随堂练兵卡》) and find out ——March / May, then learn together
  (1)Present the word, the phonetic symbol and the calendar from CAI, then Listen its pronunciation
  (2)Read the word by syllables and then read after T
  读单词的方法:小组开火车读,大小声读,男女生读
  Game: 四面开花(请一个学生起来读单词,接着此学生前后左右的学生紧跟着读。看一看哪个学生反应慢,必须再次接受处罚:独自读单词三遍。)
  (3)The same way to learn the names of the other months通过提问“哪个月份生日个数最多?”“哪两个月份名称类似”等等逐一呈现剩下的月份名并学习
  (4)Practice the words ——快速反应
  老师说节日名称,学生对出月份英语名称;老师说月份如“五月,五月”,学生说出其对应英语名称“May”;反之亦然。
  【设计意图:由针对玲玲朋友生日个数的情况提问继而学习月份名,由浅入深,逐一突破重难点。在学习新单词的过程中引导学生运用以旧带新(march—— March)、比较法(February_——January)、元音音节法来突破发音难点,模仿发音,同时结合各个月份的节日、月份的儿歌来练习记忆单词。最后让学生谈论自己喜欢的月份,进行巩固拓展。
  这种新旧知识滚雪球式的融合形式能充分调动学生积极的情感,变机械和单纯的游戏为情感的表达,回归语言学科教学的本质,让复习操练更具人文性和学科性特点。】
  3、Practice
  Play a Game:Guess Which month看由来猜月份
  Present the origins of the months by CAI
  Ss read the origins and guess the month
  CAI呈现如下。
  (1)在罗马传说中,有一位名叫雅努斯的守护神,生有先后两张脸,一张回顾过去,一张要眺望未来。人们认为选择他的名字作为除旧迎新的月名,很有意义。——January
  (2)这个月份正是大地回春,鲜花初绽的美好季节。由拉丁文“开花的日子”演变而来。——April
  【设计意图:借助外国文化知识的渗透点燃学生对月份学习的兴趣。让学生在饶有趣味的猜测中复习巩固月份单词,同时月份由来更作为课堂知识的课外延伸,达到课已尽、学无涯的目的。】
  4、Read the dialogue of Picture Two
  Read Picture Two of the lesson after the CD—Rom
  Read by themselves
  四、Extensive Reading
  1、Present the form,ask and answer
  Present the form of the birthdays from January to June in class, ask Ss to describe it
  出示一张课前老师调查的关于班级上1—6月份同学生日个数的统计表格,让学生看表格描述
  2、Practice in pairs
  Ss read the form and describe with partners
  【设计意图:语言的学习主要在于运用。在理解课文的基础上联系生活,让学生在真实自然的任务型活动中拓展运用阅读信息,补充阅读材料,让学生有话可说,提高阅读能力和语言表达能力。】
  五、Sum-up
  Sum up the lesson and expend to the emotion education——Don’t forget to express our love to parents on our birthdays. And then finish the blackboard writing.
  六、Homework
  两种不同的作业供学生根据自己的需求和能力去选择完成。
   继续完成调查表的内容,用There is/ There are 句型描述班上1-6月份生日的人数,并尝试仿写句子。如January. There are eight.
  请你运用There are句型描述1——6月份的天数,并尝试把句子写下来。

阅读更多 >>>  ignore me,Do not ignore me是什么意思?

ary怎么读

arу的发音与英语中的"are"类似,读作/ɑ?r?/,重音在第一个音节上。\n
arу是拼音字母组合,表示声母“a”和韵母“ry”组合而成,它没有固定的汉字对应,仅是用来表示外来词中的特定发音。\n
扩展知识:
拼音中的"ar"音\n"ar"是拼音中的一个音组合,一般来说,它只出现在部分外来词汇中,用来表示特定的发音。这个音组合在汉语中并没有固定的汉字对应,所以我们通常直接使用拼音字母来进行表示。
例如,在一些外来词中,我们可以看到"ar"的出现,比如"咖啡"(kā fēi)、"自行车"(zì xíng chē)等。这些词汇中的"ar"发音与英语中的"are"相似,读作/ɑ?r/。
至于为什么会有"ar"这个音组合的出现,主要是因为这些外来词在汉语中没有对应的发音组合,所以采用了这种形式来表示特定的发音。总之,"arу"这个拼音字母组合在汉语中没有固定的对应汉字,它仅用来表示特定的外来词发音。
例句:
1、Effects of Remain Content of Ti,B,Sr on Al-Si Alloys by Ary Gas Refinement氩气精炼对铝硅合金Ti、B、Sr残留量的影响
2、The Evaluation of ARY Model Spectrum Used for Observing Coloured Noise Data适用于有色观测噪声数据的AR模型谱估计
3、Carbon foams were prepared from coal tar pitch, petroleum mesophase pitch, and ary pitch.分别以煤沥青、石油中间相沥青和AR沥青为前驱体制备炭泡沫材料。

evaluation和assessment区别

evaluation和assessment区别是表达意思的不同,所属含义的不同,指代对象的不同,三个方面的区别,具体如下:
表达意思的不同:
1、evaluation是侧重于对某人或某事的价值观表现做出判断,通过评估可以确定实现目标的程度
2、assessment是评估某事物或某人的过程,即衡量质量、价值或重要性的行为,通过评估提供有关性能和改进领域的反馈
所属含义的不同:
1、evaluation是有确定规则和衡量指标的评估:
2、assessment是一种判断,评估指标来自于评估者个人,比如评估一个人的品德指代对象的不同:
1)evaluation的分析一般跟价值有关系
2)assessment指所有的通用分析,包含财务分析
evaluation和assessment造句如下:
1、The contractor's estimate for the cost of the renovation was too high.
承包商对翻新的成本估算过高。
2、The scientist's estimate of the amount of carbon dioxide emitted by the factory was based on air quality measurements and production data.
科学家对工厂排放的二氧化碳数量的估计基于空气质量测量和生产数据。
3、The evaluation of the restaurant's food and service was based on customer feedback and industry standards.
对这家餐厅的食物和服务的评价基于客户反馈和行业标准。4、The evaluation of the student's performance was based on his attendance, class participation, and test scores.
对学生表现的评价基于他的出勤情况、课堂参与度和考试成绩。

四,六级英语考试从哪年开始的?

CET阅读测试与大学英语阅读教材易读度抽样研究
辜向东 关晓仙
提 要:易读度(readability)是应用语言学中的一个重要概念,在阅读测试和阅读教材中有一定的应用价值。但目前国内对此的了解和应用并不多。本文首先简要介绍了易读度的概念,包括易读度的定义、常用的计算公式、计算步骤和它的参考量表,然后对大学英语四、六级考试的阅读测试和大学英语教材的阅读材料进行了易读度抽样研究,最后指出了目前易读度研究的局限性和今后的研究方向。
关键词:易读度,Flesh易读度公式,阅读测试,阅读教材
A Sampling Study of Readability
of CET Reading Tests and College English Reading Textbooks
Gu Xiangdong Guan Xiaoxian
Abstract: Readability is an important conception in applied linguistics and its application is of certain value in tests on reading and textbooks for reading. At present, however, it is not well-known and widely applied at home yet. This paper first gives a brief introduction to the conception, including its definition, commonly used formula, computing steps and reference table, then it makes a sampling study of the readability of reading materials in CET (Bands 4&6) and in college English textbooks, finally it points out the limitation of the present research in readability and its orientation in the future.
Key words: readability, Flesh readability formula, tests on reading, textbooks for reading
CET阅读测试和大学英语阅读教材易读度抽样研究[本文为重庆大学基础及应用基础研究基金项目“语言测试与语言教学”研究成果之一。
辜向东(1967-)女,重庆大学副教授,现上海交通大学博士生,研究方向语言测试。
关晓仙(1977-)女,上海交通大学硕博连读生,研究方向语言测试。]
辜向东 关晓仙
(上海交通大学外国语学院, 上海 200240)
1. 引言
易读度(readability),也称易读性或可读性,是应用语言学中的一个重要概念,在阅读测试和阅读教材的选材、命题及评估中都有一定的应用价值。但目前国内对此的了解和应用并不多。本文首先简要介绍了易读度的概念,包括易读度的定义、常用的计算公式、计算步骤和它的参考量表,然后对大学英语四、六级考试的阅读测试和大学英语教材的阅读材料进行了易读度抽样研究,最后指出了目前易读度研究的局限性和今后的研究方向。
2. 易读度研究
2.1易读度的概念
《朗文语言教学及应用语言学辞典》(Longman Dictionary of Language Teaching & Applied Linguistics) 对易读度的定义是:
How easily written materials can be read and understood. Readability depends on many factors, including (a) the average length of sentences in a passage, (b) the number of new words a passage contains, (c) the grammatical complexity of the language used. Procedures used for measuring readability are known as “readability formulae” (Richards, etc., 1998:383-384).
因此,“易读度”,顾名思义,容易阅读的程度。影响易读度的因素主要是语言的形式,如句长、生词量、语法复杂程度等。迄今为止,国际上从事阅读研究的应用语言学家和心理语言学家提出了不少计算易读度的公式。其中,比较常用的有Flesh易读度公式、Lorge易读度公式和Bormuth阅读能力等级公式等。这些公式都力求从语言形式上对阅读材料的易读度进行量化。以下介绍一个简单实用的计算公式,“Flesh易读度公式”(Flesh Readability Formula)。
Readability = 206.84 – 0.85 wl – 1.02 sl
其中: wl 指每100个单词中的音节数 (the number of syllables per 100 words);
sl 指每个句子的平均词数(the average number of words per sentence)。
(杨惠中,1997:67)
易读度的计算分为四个步骤:
第一步:统计出每篇阅读材料的总词数、总音节数、总句子数;
第二步:用总音节数除以总词数,计算出每个词的平均音节数,乘以100得出每100
词的音节数;
第三步:用总词数除以总句子数,计算出每个句子的平均词数;
第四步:将每100词的音节数和每个句子的平均词数代入Flesh易读度公式,计算出每
篇阅读材料的易读度。
易读度的参考量表分为七个等级,从非常难到非常容易,易读度越低的材料阅读难度越大。如:易读度在0-30的材料非常难,有这样难度的典型杂志是 “Scientific”,读者达到大学水平,具有阅读这种材料能力的人在美国成年人中的比例为4.5%;易读度在90-100的材料非常容易,这一类的典型杂志是 “Comics”,读者为小学四年级水平,具有阅读这种材料能力的人在美国成年人中的比例为90%。
Readability
score Description of style Typical magazine Potential readers
School grade % of US Adults
0-30
30-50
50-60
60-70
70-80
80-90
90-100 Very difficult
Difficult
Fairly difficult
Standard
Fairly easy
Easy
Very easy Scientific
Academic
Quality
Digest
Slick-fiction
Pulf-fiction
Comics College
High school
Some H. S.
7 or 8 th grade
6 th grade
5 th grade
4 th grade 4.5
24
40
75
80
86
90
表一 易读度参考量表 (Reference table of Readability)
(转引自杨惠中博士学位课程“语言测试理论与实践”讲义,2001)
2.2 易读度原则
虽然易读度公式只考虑阅读材料的形式要素,而且是为以英语为母语的人设计的,对于以英语为外语的成年人在阅读中的许多困难没有考虑进去,但它为我们阅读测试和阅读教材材料的选择和评估提供了一个重要的参考依据。
目前我国的英语测试和英语教学主要采用分级制,如:大学英语考试分为四级和六级,大学英语教学分为基础阶段和应用提高阶段。分级制的特点是级别越高,难度越大。根据易读度的概念,在阅读测试与阅读教材中,级别越高,材料的易读度越低,而且同级别的阅读测试或教材其材料的易读度应该基本一致。笔者将此称为“易读度原则”。本文对我国的大学英语阅读测试与阅读教材是否遵循了这一原则进行了抽样研究。
2.3 大学英语四、六级考试阅读测试材料易读度研究
《大学英语四、六级考试效度研究》曾经做过1987-1995年历次阅读测试材料的易读度研究。研究结果表明,六级阅读材料的平均易读度(49.1)明显低于四级阅读材料的平均易读度 (57.7)。根据易读度参考量表,六级属于 “difficult” (30-50)的等级,四级属于 “fairly difficult”(50-60)的等级(杨惠中,1997:67-68)。
1999年《大学英语教学大纲》[修订本]的颁布与实施标志着大学英语教学进入了一个新的阶段。为了检验四、六级阅读测试材料的易读度是否体现了《大纲》的更高要求,本文对《大纲》[修订本]执行后最新公布的四、六级考试试题,即2000年6月考次四级和六级的八篇阅读测试材料进行了易读度研究(见表二)。
材料
级别 Passage 1 Passage 2 Passage 3 Passage 4 均值
四级 44.90 48.00 50.11 55.86 49.72
六级 43.54 44.83 31.32 34.52 38.54
表二 2000年6月考次四级和六级阅读测试材料的易读度
根据表二的易读度统计,大学英语六级每篇阅读材料的易读度均低于四级每篇阅读材料的易读度,而且四级和六级的易读度没有丝毫重叠,四级的上限为44.9,六级的下限为44.83。这样四级和六级的难度被区别开来了。这在一定程度上说明了大学英语四、六级考试命题的科学性。
2000年6月考次四级和六级阅读测试材料易读度的均值分别为49.72和38.54,而1995年6月之前历次考试(四级共15次,六级共13次)四级和六级阅读材料的易读度的均值分别为57.1和49.1。很显然,现在的大学英语四、六级阅读测试材料的难度有大幅度的增长,四级的难度接近过去六级的难度,六级的难度也有相应地增加。如果我们比较2000年6月、1995年6月和1990年6月三次四、六级考试阅读材料的易读度的均值(见表三),我们会发现这种增长与十年间两次大学英语教学的重大改革密切相关。第一次是1994年大学英语文理科和理工科两个教学大纲词表合二为一,1995年在考试中实行。在此之前,四、六级考试是按两个大纲进行的。由于两个大纲四级词表共同部分只占80%,因此实际考试的四
级别、时间
篇章、变量 四级 六级
1990.6 1995.6 2000.6 1990.6 1995.6 2000.6
P1 61.3 49.4 44.9 48.9 53.0 43.54
P2 72.2 61.3 48.0 47.0 40.3 44.83
P3 42.1 46.8 50.1 27.5 28.4 31.32
P4 62.0 53.6 55.9 55.0 35.6 34.52
均值 59.5 52.8 49.7 44.5 39.3 38.54
跨越参考量表等级数 4 3 2 3 3 1
波动幅度 30.1 14.5 11 27.5 24.6 13.51
表三 1990.6、1995.6、2000.6 三次考试阅读材料易读度对比
级词汇量只有3200词左右,六级词汇量也未达到大纲规定的5300词的要求。在四、六级阅读测试中体现为阅读材料的易读度较高,1990年6月四级平均为59.5,六级平均为44.5,即阅读测试材料的难度较低。1995年开始实施大学英语通用词汇表四级4000词,教学要求上了一个新的台阶。这在大学英语四、六级阅读测试材料的易读度中即刻得到了体现。1995年6月四、六级阅读测试材料易读度的均值明显下降,分别降至52.8和39.3,即阅读测试材料的难度明显加大。大学英语教学的第二次重大改革是1999年《大纲》[修订本]的颁布和实施。由于四、六级阅读测试材料的难度已经较高,所以2000年6月考次四级和六级阅读测试材料的易读度均值没有1995年6月考次下降的幅度那么大,但也有所下降,分别降至49.7和38.54。也就是说,阅读测试材料的难度也有一定的增加。
按照理想的模式,属于同一级的所有阅读测试材料的难度应该处于同一层次,如四级均在“fairly difficult”(50~60)之间;不同级别的阅读测试材料其难度应该有明显区分,不能重叠,如四级在“fairly difficult”区间内(即易读度在50-60之间),六级在“difficult” 区间内(即易读度在30-50之间)。表四“理想的阅读测试模式与实际的分布之间的对照”和表三“跨越参考量表等级数”和“波动幅度”栏目表明,大学英语四、六级阅读测试随着实践的增加,在不断趋于合理与完善。1990年6月考次四级阅读材料的易读度跨越了四个等级,且偏高,波动幅度较大;六级三个等级,其中一篇的易读度偏低,波动幅度也较大,整体的难度偏低。1995年6月考次四级跨越的等级降为三个,六级仍然为三个,但波动幅度明显下降。2000年6月考次四级阅读材料的易读度集中在两个等级,六级在一个等级,两者波动的幅度均较小,十分接近理想的模式。
易读度区间
难度描述
理想
的模式
实际的分布
1990.6 1995.6 2000.6
四级 六级 四级 六级 四级 六级
0~30 Very difficult P3 P3
30~50
Difficult
六级
P3 P1,
P2 P1,
P3 P2,
P4 P1,
P2 P1, P2, P3, P4
50~60 Fairly difficult
四级
P1
P4
P4
P1 P3
P4
60~70
Standard
P4
P2
70~80 Fairly easy
P2
80~90 Easy
90~100 Very easy
表四 理想的阅读测试模式与实际的分布之间的对照
2.4大学英语阅读教材材料易读度研究
众所周知,阅读是《大纲》第一层次的要求,而“教材是实现教学大纲确定的教学目标的重要保证”(《大纲》,1999:12)。现行的大学英语教材在全国使用范围最广、影响最大的有四套:《大学英语》、《21世纪大学英语》、《新英语教程》和《新编大学英语》。这些教材都是根据《大纲》[修订本]的要求修订或新编写的。因为时间、精力和条件的限制,本文只选择了其中一套教材进行易读度抽样研究。该教材共4册,每册10个单元,每个单元由同一题材的三篇文章组成。课文A为精读材料,B、C为泛读材料。本文抽取了该教材每一册的第一单元与第十单元,共24篇阅读课文。这样抽样的目的是为了做一系列的比较研究:(1)比较精读和泛读材料的易读度;(2)比较同一册阅读材料的易读度;(3)比较不同册的阅读材料的易读度。表五为24篇阅读课文的易读度统计。
材 料
册数/单元 A B C 均值 跨越参考量表等级数 波动
幅度
第一册 Unit 1 56.68 69.25 94.28 73.40 67.42 5 37.60
Unit 10 65.26 57.71 61.35 61.44
第二册 Unit 1 66.37 60.54 42.36 56.42 53.09 3 24.01
Unit 10 46.07 44.99 58.18 49.75
第三册 Unit 1 74.11 73.69 62.05 69.95 68.64 2 15.72
Unit 10 65.15 60.55 76.27 67.32
第四册 Unit 1 56.45 67.15 50.35 57.98 61.54 3 22.04
Unit 10 72.39 63.31 59.6 65.10
平均 62.81 62.15 63.06 62.67 3.25 24.84
表五 某大学英语教材24篇阅读材料的易读度统计
从总体(即均值)上看,该教材精读材料(A)和泛读材料 (B、C) 的易读度基本一致,分别为62.81,62.15,63.06。说明精读和泛读材料的难度总体上没有大的差异。除第一册外,每册跨越参考量表的等级数在2-3之间,每册六篇阅读材料易读度波动的幅度不是太大。这说明同一册阅读材料的难度比较一致。册与册之间的易读度均值呈波浪状起伏,总趋势是册数越高,易读度越低。但除第二册外,册与册之间易读度的区分度不够明显。
根据分级教学的特点,理想的阅读教材的模式为(1)精读材料的难度略高于泛读材料;
(2)同一册的阅读材料的难度应该处于同一层次,如第一册所有的阅读材料均在“fairly easy”(70~80)之间;(3)不同册的阅读材料其难度应该有明显区分,如第二册在“standard”(60-70)之间。不过教材的特点是由浅入深,螺旋式上升,因此相连册的阅读材料难度的等级容许部分重叠。
因为本文研究的教材是我国在校大学生使用的,他们入学前已经有一定的阅读能力,所以在理想的模式中,阅读材料的易读度不应该包括80~90和90~100两个等级,即阅读材料的难度不应该非常容易和容易。与此同时,1-4册属于基础阶段的教材,其阅读材料的易读度也不应该包括0~30等级,即其难度不应该非常难。
根据表六“理想的阅读教材模式与实际分布的对照”,第一册1-1-C太简单,不宜使用;1-1-A和1-10-B却比较难,适合作为第三册的材料;第二册有三篇材料较难,2-1-C, 2-10-A和2-10-B,适合作为第四册的材料;第三册整体偏易,适合作为第一或第二册的材料;第四册有三篇材料偏易,4-10-A,4-10-B和4-1-B,也适合作为第一或第二册的材料。
易读度区间 难度描述 理想的模式 实际的分布
0~30 Very difficult
30~50 Difficult 第四册 2-1-C,2-10-A,2-10-B
50~60
Fairly difficult
第三册 1-1-A,1-10-B,2-10-C
4-1-A,4-1-C,4-10-C
60~70
Standard
第二册 1-1-B,1-10-A,1-10-C,2-1-B,2-1-C
3-10-A,3-10-B,3-1-C,4-10-B,4-1-B
70~80 Fairly easy 第一册 4-10-A,3-1-A,3-1-B,3-10-C
80~90 Easy
90~100 Very easy 1-1-C
表六 理想的阅读教材模式与实际的分布之间的对照
(注:理想的模式中箭头表示相连册的难度的等级容许部分重叠;
实际的分布中第一个数字表示册数,第二个数字表示单元)。
3. 结语
本文对易读度概念的介绍和对大学英语阅读测试及阅读教材材料易读度的抽样研究意在帮助人们了解易读度的应用价值并在相关领域给予足够的重视。本研究结果表明,大学英语四、六级阅读测试较好地遵循了易读度原则,而大学英语阅读教材在此方面有必要做些改进。不过,本文运用的易读度公式只考虑了语言材料的形式要素,如词长和句长,而且该公式是为以英语作母语的人设计的,其参考量表也是以英语为母语的人作参照的,对于以英语为外语的中国学生在阅读中的许多困难没有考虑进去。事实上,除易读度外,阅读测试和阅读教材材料的选择和评估还要考虑很多别的因素,如:语言的真实性(authenticity)、体裁的多样性(variety)、信息的密度(density)等。此外,本文的抽样量比较有限。所以,该研究结果只能作为大学英语阅读测试和阅读教材选材以及评估的一个参考。今后易读度研究的方向是设计出既考虑形式也考虑语义的易读度公式,计算出符合中国学生实际的易读度参考量表,并在阅读测试与阅读教材中广泛运用。
参考文献
Alderson, JC. Caroline Clapham and Dianne Wall. Language Test Construction and Evaluation [M]. Shanghai: Foreign Language Teaching and Research Press, 2000.
Richard, J., John Platt and Heidi Weber. Longman Dictionary of Language Teaching & Applied Linguistics [M]. Essex: Longman Group Limited, 1985.
Yang, HZ, A new technique for identifying scientific/technical terms and describing science texts [J]. Literary and Linguistic Computing. 1985, (1): 93-103.
大学英语教学大纲词表调整工作组,大学英语教学大纲通用词汇表(1-4级)[M]。上海:上海外语教育出版社;北京:高等教育出版社,1994。
大学英语教学大纲词表修订工作组,大学英语教学大纲词汇表 [M]。上海:上海外语教育出版社;北京:高等教育出版社,2000。
大学英语教学大纲修订工作组,大学英语教学大纲 [修订本] [M]。上海:上海外语教育出版社;北京:高等教育出版社,1999。
复旦大学,上海交通大学,21世纪大学英语读写教程1-4册 [M]。上海:复旦大学出版社;北京:高等教育出版社,1999。
黄人杰,杨惠中,从统计角度分析科技英语词汇 [J]。外语教学与研究,1985,(4):34-39。
黄人杰,杨惠中,计算机辅助科技英语词汇统计结果的初步分析 [J]。外国语,1984,(1):46-51。
全国大学英语四、六级考试委员会,大学英语六级考试历年实考卷与详解 [M]。上海:上海外语教育出版社,上海外语音响出版社,2001。
全国大学英语四、六级考试委员会,大学英语四级考试历年实考卷与详解 [M]。上海:上海外语教育出版社,上海外语音响出版社,2001。
上海外国语大学TEM考试中心,TEM考试效度研究 [M]。上海:上海外语教育出版社,1997。
杨惠中和C. Weir,大学英语四、六级考试效度研究 [M]。上海:上海外语教育出版社,1998。
杨惠中,博士学位课程“语言测试理论与实践”讲义,上海:上海交通大学,2001。作者联系方式:
通讯地址:(200240)上海市东川路800号上海交通大学A0114021信箱
电子邮件:辜向东:xiangdonggu@263.net
关晓仙:stella_gxx@sina.com
电话号码:辜向东:021-54741271; 023-65102554
关晓仙:021-54741947

阅读更多 >>>  makefile语法,makefile怎么用?

网站数据信息

"evaluation有几个音节,Evaluationi,væljʊ'eʃən 这个音标里逗号和上引号分别代表什么?请英语大神解答~"浏览人数已经达到19次,如你需要查询该站的相关权重信息,可以点击进入"Chinaz数据" 查询。更多网站价值评估因素如:evaluation有几个音节,Evaluationi,væljʊ'eʃən 这个音标里逗号和上引号分别代表什么?请英语大神解答~的访问速度、搜索引擎收录以及索引量、用户体验等。 要评估一个站的价值,最主要还是需要根据您自身的需求,如网站IP、PV、跳出率等!