同等学力英语每日一练:2016年12月28日
Computers can beat chess champion Gary Kasparov at his game,count all the atoms in a nuclear explosion,and calculate complex figures in a fraction of a second,but they still fail at the slight differences in language translation. Artificial Intelligence computers have large amounts of memory,capable of storing huge translating dictionaries and extensive lists of grammar rules. Yet,today’s best computer language translators have just a 60 percent accuracy rate. Scientists are still unable to program the computer with human-like common sense reasoning power.
Computer language translation is called Machine Translation,or MT. While not perfect,MT is surprisingly good. MT was designed to process dry,technical language that people find tedious to translate. Computers can translate basic phrases,such as“You foot bone’s connected to your ankle bone,your ankle bone’s connected to your leg bone.”They can translate more difficult phrases,such as“Which witch is which?”Computers can also accurately translate“Wild thing,you make my heart sing!”into other languages because they can understand individual words,as long as the words are pre-programmed in their dictionary.
But highly sensitive types of translating,such as important diplomatic conversations,are beyond the scope of computer translating programs. Human translators use intuitional meaning,not logic,to process words and phrases into other languages. A human can properly translate the phrase,“The pen is in the pen(围养禽畜的圈),”because most humans know that it means that a writing instrument is in a small enclosed space. Many times,computers do not have the ability to determine in which way two identical words in one sentence are to be used.
In addition to using massive rule-programmed machines,computer programmers are also trying to teach computers to learn how to think for themselves through the“experience”of translating. Even with these efforts,programmers admit that a“thinking”computer might not ever be invented in the future.
46. Computers today are capable of ( ).
A. defeating the best chess player in the world
B. telling subtle differences between languages
C. translating over 60 percent of difficult texts
D. doing human-like common sense reasoning
47. According to the passage,which of the following is NOT true?
A. Computers can translate dry and difficult phrases.
B. Computers can understand sensitive language.
C. Computers can translate technical language.
D. Computers can understand pre-programmed words.
48. The major problem with computer translating programs is that computers ( )
A. can not translate illogical sentences
B. do not have a large enough capacity of memory
C. can not understand grammatical rules
D. do not have intuition to process language
49. To improve machine translation,computer programmers are trying to ( ).
A. use powerful rule-programmed computers
B. teach computers to think by practice
C. have computers compile translating dictionaries
D. add explanations of words in computer programs
50. The passage suggests that ( ).
A. the accuracy rate of machine translation cannot be raised
B. it is impossible for computers to think as humans do
C. only technical language is suitable for machine translation
D. it is impossible to determine of identical words
译文:
尽管计算机能够打败围棋冠军加里·卡斯帕洛夫,计算出在一次核爆炸中的原子数,几秒钟得出复杂的计算结果,但它们还是在语言翻译的细微区别上无能为力。人工智能的计算机有着巨大的存储能力,能存储庞大的翻译字典和广阔的语法规则。然而,现今最好的计算机语言翻译器都只能达到60%的正确率。科学家们至今还无法编辑出具有和人类相当逻辑推理能力的计算机。
计算机语言翻译器又被称为“翻译机器”,或MT。尽管还不很完善,翻译机器还是相当令人惊讶。它是用来处理枯燥的技术性语言,人们通常认为这过于乏味。计算机能翻译基本的短语,比如“你的脚骨和你的踝关节相连,你的踝关节和你的腿相连”。它们还能翻译更复杂点的短语,比如“那个巫师到底是谁?”计算机还能准确地把“大自然,你使我的心高歌”翻译成其他语言,因为它们只能理解单个的词语,只要将这些词语预先在程序里设定。
但涉及敏感话题的翻译,诸如重要的外交语言,往往超出计算机翻译程序的范围。人类翻译靠的是直觉而不是逻辑,来将词汇和短语处理成其他语言。人脑能准确地翻译“钢笔在猪圈里”,因为大部分人知道这实际上是指书写工具困于狭窄的空间。而很多时候,计算机不能决定相同的词语在同一句子里如何被使用。
除了使用强大的语法规则翻译机器之外,程序师还同时教计算机如何去通过翻译的经验自己独立思考。即使基于这些努力,程序师承认有思考能力的计算机恐怕将永远不会被发明创造。
答案解析:
46.【正确答案】A
【考点类型】细节判断
【解析过程】今天计算机可以能够
A. 打败世界上最好的象棋选手
B. 区分语言间的细微差别
C. 翻译60%的疑难的文章
D. 从事类似人类的常识推理
线索:“Computers can beat chess champion Gary Kasparov at his game,”表明选项A是正确答案。B选项和原文“but they still fail at the slight differences in language translation.”截然相反。C选项相同,和原文内容“Scientists are still unable to program the computer with human-like common sense reasoning power.”相反。
【考点提示】文章的各段首句一般都有可能成为考试重点,比如同义词替换或正话反说等都是出题老师有可能设计答案的方式。
47.【正确答案】B
【考点类型】细节判断
【解析过程】根据文章,下面哪句话是错误的?
A. 计算机能够翻译枯燥难解的短语
B. 计算机能够理解敏感的语言
C. 计算机能够翻译技术用语
D. 计算机能够理解预先在程序中设定的词汇
线索:“But highly sensitive types of translating,such as important diplomatic conversations,are beyond the scope of computer translating programs.”表达得很清楚,目前计算机是无法完成处理如外交词汇这类比较敏感的术语。选项A、C、D在文章的第2段都明确地给出了它们的出处。
【考点提示】三对一错这类细节判断题是比较难的一种题型,对于这种题型,同学们首先要明确考试的内容有可能出现的地方,一般是可以根据题号判断,如第2题有可能在2/3段出现。其次,也可以根据文章中的一些线索词判断,如表示转折的词,表示因果关系的词,表示总结或绝对化表达的词等。
48. 【正确答案】A
【考点类型】细节判断
【解析过程】计算机翻译程序的主要问题是计算机
A. 不能翻译不符合逻辑的语言
B. 没有大量的记忆能力
C. 不能够理解语法规则
D. 没有处理语言的直觉
线索:文章第3段的后部“Many times,computers do not have the ability to determine in which way two identical words in one sentence are to be used.”表明计算机没有用直觉来处理语言的能力。选项B肯定是错误的,原因是它首先就和我们的常识不符合,同时和文章第1段的内容“Artificial Intelligence computers have large amounts of memory”也相反。D选项是偷换概念,原文根本就没有提到计算机有没有直觉和处理语言直接有何关系。
【考点提示】有些混淆选项特别容易判断出来,原因是这些选项的设计比较有规律可以遵循,如选项B,它很明显和我们的常识就不符合,计算机的存储能力应该是不成问题的。
49.【正确答案】B
【考点类型】细节判断
【解析过程】为了提高计算机翻译的能力,程序设计人员正在试图
A. 使用强大的语法规则翻译计算机
B. 教计算机用实践经验去思考
C. 让计算机汇编翻译词典
D. 在计算机的程序中加入单词解释
线索:“In addition to using massive rule-programmed machines,computer programmers are also trying to teach computers to learn how to think for themselves through the‘experience’of translating.”表明选项B是正确答案。In addition to表示除了的含义,重点应该放在它的后面主句,所以A不正确。
【考点提示】文章中常会出现一些表示逻辑关系的连接词,比如转折关系,递进关系,因果关系,对比关系等,这些连接词需要同学们自己多总结,有很强的敏感度,在考试的时候,一看到这类词汇,就能马上知道这是考点。
50.【正确答案】B
【考点类型】主旨推断
【解析过程】文章暗示:
A. 机器翻译的准确率不可能提高
B. 计算机不可能像人类一样去思考
C. 仅仅技术语言是适合计算机翻译
D. 确定含义相同的词汇是不可能的
线索:文章最后一段指出“ Even with these efforts,programmers admit that a‘thinking’computer might not ever be invented in the future.”表明选项B是正确答案。这里,同学们要注意的是作者用的情态动词might not ever be表示很不可能发生。C选项中only表示绝对化的含义,一般是不正确的答案。
【考点提示】选项中如果出现表示绝对化的词如:only,first,most,least,best 等信息一般都是不正确的答案,这一点是和出题老师设计试题一定要坚持严谨的原则有关。