Abstract Toward Memory--based Translation
2021年安徽省中考英语考纲词汇表
A America balloon boreda(an)American bamboo boringability among banana bornable ancient bank borrowabout and baseball bossabove angry basic bothabroad animal basket bottleabsent another basketball bottomaccept answet bathroom bowlaccident ant be(am,is,are)boxaccording to any beach boyachieve anybody bean brainanyone bear braveact anything beat breadaction anyway beautiful breakactive anywhere because breakfastactivity appear become breathactor apple bed bridgeactress April bedroom brightadd area beef bringaddress arm before Britainadult army begin British advantage around behind brotheradvice arrive believe brownadvise art bell brushafford article belong buildafraid artist below buildingAfrica as beside burnAfrican Asia best busafter Asian better business afternoon ask between busyagain asleep big butagainst at bike(=bicycle)butterage attend bill buyago attention bird byagree August birth Cagreement aunt birthday cakeahead Australia biscuit calendarair Australian bit callairport autumn black cameraalive avoid blackboard campall awake blind can(can't/cannot) allow away block Canadaalmost awful blood Canadianalone B blouse cancelalong baby blow canceraloud back blue candle already background board candy also bad boat cap although bag body capital always ball book carcard club create discuss care coach cross discussion careful coal cry dish careless coast culture divide carrot coat cup docarry coffee customer doctor cartoon coin cut dogcat cold dollar catch collect D doorcause college daily double celebrate colour(AmE color)dance doubtcent come danger downcentral comfortable dangerous drawcentre(AmE centet)common dark dream century communicate date dress certain communication daughter drink certainly community day drivechair company dead driver chalk compare deaf drop chance competition deal drychange complete dear duckcheap computer December dumpling check concert decide during cheer condition decision duty chemistry connect deep E chess consider degree each chicken continue delicious earchild(pl,children)control dentist earlyChina conversation depend earth Chinese cook describe earthquake chocolate cookie design eastchoice cool desk eastern choose copy develop easy chopsticks corn development eat Christmas corner dialogue(AmE dialog)education church correct diary effort cinema cost dictionary eggcircle cotton die eightcity cough difference eighteen class could different eighth classmate count difficult eighty classroom country difficulty either clean countryside dig elderclear couple dining electronic clever courage dinner elephant climb course direct elevenclock cousin direction else close cover director email clothes cow dirty empty cloud crayon discover encourage cloudy crazy discovery endenergy farmer fourth greeting enemy fast France grey engineer fatfree ground England father(dad)French group English favourite(AmE favorite)fresh grow enjoy fearFridayguard enough February fridge(=refrigerator )guess enterfeed friend guest environment feel friendly guide eraser feeling friendship guitar especially festival from gunEurope fever front HEuropean few fruit habit even field full hair evening fifteen fun half event fifth funny hallever fifty futurehamburger every fight Ghand everybody fill game handbag everyday film garden handsome everyone finally gate hang everything find general happen everywhere fine gentleman happy exactlyfinger geography hard exam(=examination)finish German hardly example fire Germany harmful excellent first get hat except fishgift hate excited fisherman giraffe have(has)exciting fit girl he excuse five give head exercise fix glad headache expect flag glass health expensive floor glove healthy experience flower glue hear explain fly go heart express flu gold heat eyefollow good heavy Ffoodgoodbye(bye)height face foot(pl. feet)government hello fact football grade help factory for grammar helpful fail force granddaughter hen fair foreign grandfather(grandpa)her fall forest grandmother(grandma)here familyforgetgrandparent herofamous fork grandson hers fan form grape herself fantastic forty grass hifar four great hide farm fourteen green highhill infet knife(pl.knives)longhim influence knock lookhimself information know losehis inside knowledge lothistory instead L loudhit instruction lab(=laboratory)lovehobby instrument lady lovelyhold interest lake lowhole interesting land luckholiday international language luckyhome Internet large lunchhometown intetview last M homework into late machinehonest introduce later madhonour(AmE honor)introduction laugh magazinehope invent law magichorse invention lay mainhospital invite lazy makehot island lead man(pl.men)hotel it leader managehour its leaf(pl.leaves)managerhouse itself learn manyhousework J least maphow jacket leave Marchhowever January left markhuge Japan leg markethuman Japanese lemon marryhumorous job lend matchhundred join less maths(=mathematics) hungry joke lesson matterhurry journey let mayhurt juice letter Mayhusband July level maybeI jump library meI June lie mealice just life meanice-cream K lift meaningidea keep light meatif key like medicalill keyboard line medicineillness kick lion meetimagine kid list meetingimportant kill listen member impossible kilo(=kilogram)litter memoryimprove kilometre(AmE kilometer)little mentionin kind live menuinclude king lively messincrease kiss local messageIndia kitchen lock methodIndian kite London metre(AmE meter) industry knee lonely middlemight neithet one personal mile nervous online pet milk never only phonemind new open photo(=photograph)mine newsopinion physics minute newspaper opposite piano mirror next or pick Miss nice orange picnic miss night order picture missing nine other pie mistake nineteen our piece mixninety ourspig mobile phone ninth ourselves pilot model no out pink moderm nobody outside pioneer moment nod over pity Monday noise ownplace money noisy Pplan monkey none Pacific plane month noodle page planet moon noon pain plant more nor paint play morning north pair playermostnorthern palace playground mother(mum)nose pale please mountainnot pancake pleasure mouse(pl.mice)note panda plenty mouth notebook paper pocket move nothing paragraph poem movie notice pardon point Mr Novembet parent policeMrs now park policeman/policewoma n Ms number part polite much nutseparty pollute museum Opass pollution music object passage pool musician ocean passenger poormust o'clock passport popular(pop)my October past population myselfof patient possible Noff paypost name offer PE(=physical education)postcard national office peace postman natural officer pear potato nature often pen pound nearoilpencil powernearly OK people practice necessary old perfect praise neck Olympics perhaps prefer need on period prepare neighbour(AmE neighbor)once person presentpresident really sad shepretty reason safe sheep(pl.sheep) price receive safety shinepride recently salad shipprimary record sale shirtprint red salt shoeprivate refuse same shopprize regard sand short probably regret sandwich shorts problem relationship satisfy should produce relative Saturday shoulder product relax save shout programme(AmE program)remain say show progress remember scarf shower project repair school shutpromise repeat schoolbag shy pronounce reply science sick pronunciation report scientist sideproper require scissors sightprotect research score signproud respect screen silence provide rest sea silentpublic restaurant search silkpull result season sillypunish return seat silverpupil review second similarpurple rice secret simple purpose rich secretary sincepush ride see singput right seem single Q ring seldom sirquarter rise sell sisterqueen risk send sitquestion river sense situation quick road sentence sixquiet robot separate sixteenquite rock September sixthR role serious sixtyrabbit room serve sizerace rope service skateradio rose set skillrailway round seven skirtrain row seventeen skyrainy rubbish seventh sleepraise rule seventy sleepyrapid ruler several slowreach run shake small read rush shall smart ready Russia shame smell real Russian shape smile realise(AmE realize)S share smokesmooth still task tiesnake stomach taste tigersnow stomachache taxi timesnowy stone tea tinyso stop teach tiredsocial store teacher titlesociety storm team tosock story technology todaysofa straight teenager togethersoft strange telephone toiletsolve stranger television(=TV)tomato some strawberry tell tomorrow somebody street temperature ton someone strict ten tonight something strong tennis too sometimes student tenth tool somewhere study term tooth(pl.teeth) son stupid terrible toothache song subject test topsoon succeed text totalsorry success than touchsound successful thahk toursoup such that touristsouth sudden the toward(s) southern sugar theatre(AmE theater)towerspace suggest their townspare suggestion them toyspeak summer themselves trade speaker sun then traditional special Sunday there traffic speech sunny these trainspeed supermarket they trainingspell supper thick translate spend support thin travelspirit suppose thing treasure spoon sure think treatsport surface third treespread surprise thirsty tripspring survey thirteen trouble square sweater thirty trousers stamp sweet this truckstand swim those TRUE standard swimming though truststar T thought truthstart table thousand trystate tail three T-shirtstation take through Tuesday stay talent throw turn steal talk Thursday twelfth step tall ticket twelve stick tape tidy twentiethtwenty waste withtwice watch withouttwo water wolfU watermelon woman(pl.women) ugly way wonderUK we wonderfulumbrella weak wooduncle wealth wordunder wear workunderground weather workerunderlined website worldunit Wednesday worryunderstand week worseuniversity weekday worstunless weekend worthuntil(till)weigh wouldup weight woundupon welcome writeus well writerUS west wronguse western Xused wet X-rayuseful what Yusual whatever yardusually wheel yearV when yellowvacation whenever yesvaluable where yesterdayvalue whether yetvegetable which youvery while youngvictory white yourvideo who yoursvillage whole yourself(pf.yourselves) violin whom Zvisit whose zebravisitor why zerovoice wide zoovolleyball wifevolunteer wildW willwait winwake windwalk windowwall windywallet winewant wing war winner warm winter warn wise wash wish。
空间基因驻留地方记忆的规划设计新途径——南京原近代民国首都机场案例
空间基因：驻留地方记忆的规划设计新途径—南京原近代民国首都机场案例□ 邵润青，段 进，钱 艳，王里漾，姜 莹[摘 要]针对在城市片区发展变迁中地方记忆驻留的难题，文章提出了基于空间基因的规划设计新途径，借助空间基因的地方信息承载作用，实现了地方记忆的驻留。
首先，解析获取承载地方特征信息的空间基因，同时结合深层结构的发展变化情况，判断其传承与发展的必要性与可能性。
其次，将筛选出的空间基因及其特征因子传承在满足发展需求的新的空间结构中，并通过设计演绎，再现出具有历史特征记忆的新场景，从而在规划设计中完成历史信息传递，留住令人留恋的地方记忆。
该方法在南京原近代民国首都机场片区搬迁再开发项目中得到了成功应用。
[关键词]地方记忆；空间基因；规划设计方法[文章编号]1006-0022(2020)19-0040-07 [中图分类号]TU984 [文献标识码]B[引文格式]邵润青，段进，钱艳，等．空间基因：驻留地方记忆的规划设计新途径—南京原近代民国首都机场案例[J]．规划师，2020(19)：40-46．Space Gene: A New Approach of Planning and Design to Promote the Sense of Place: T aking NanjingDajiaochang Airport As an Example/Shao Runqing, Duan Jin, Qian Y an, Wang Liyang, Jiang Ying[Abstract]Aiming at the problem of local memory retention in the development and change of urban area, a new approach ofplanning and design based on space gene is proposed. It preserves the local sense with the space gene’s local informationcarrying functions. First, space genes are analyzed. Second, the necessity and possibility of space genes’ continuation aredetermined by the coupling with local development goals. Finally, through design strategies, space genes are anchored in theurban structure with new development demands. Thus the reproduced local identity will arouse public memory through thehistorical information transmission. The approach has been successfully applied in the relocation and redevelopment project of theNanjing Dajiaochang Airport area.[Key words] Local memory, Space gene, Planning and design approach[基金项目] [作者简介]国家重点研发计划(2019YFD1100700)邵润青，博士，高级工程师，东南大学城市规划设计研究院副院长。
模糊逻辑中英文对照外文翻译文献
模糊逻辑中英文对照外文翻译文献(文档含英文原文和中文翻译)译文:模糊逻辑欢迎进入模糊逻辑的精彩世界，你可以用新科学有力地实现一些东西。
在你的技术与管理技能的领域中，增加了基于模糊逻辑分析和控制的能力，你就可以实现除此之外的其他人与物无法做到的事情。
以下就是模糊逻辑的基础知识：随着系统复杂性的增加，对系统精确的阐述变得越来越难，最终变得无法阐述。
于是，终于到达了一个只有靠人类发明的模糊逻辑才能解决的复杂程度。
模糊逻辑用于系统的分析和控制设计，因为它可以缩短工程发展的时间；有时，在一些高度复杂的系统中，这是唯一可以解决问题的方法。
虽然，我们经常认为控制是和控制一个物理系统有关系的，但是，扎德博士最初设计这个概念的时候本意并非如此。
实际上，模糊逻辑适用于生物，经济，市场营销和其他大而复杂的系统。
模糊这个词最早出现在扎德博士于1962年在一个工程学权威刊物上发表论文中。
1963年，扎德博士成为加州大学伯克利分校电气工程学院院长。
那就意味着达到了电气工程领域的顶尖。
扎德博士认为模糊控制是那时的热点，不是以后的热点，更不应该受到轻视。
目前已经有了成千上万基于模糊逻辑的产品，从聚焦照相机到可以根据衣服脏度自我控制洗涤方式的洗衣机等。
如果你在美国，你会很容易找到基于模糊的系统。
想一想，当通用汽车告诉大众，她生产的汽车其反刹车是根据模糊逻辑而造成的时候，那会对其销售造成多么大的影响。
以下的章节包括：1)介绍处于商业等各个领域的人们他们如果从模糊逻辑演变而来的利益中得到好处，以及帮助大家理解模糊逻辑是怎么工作的。
2)提供模糊逻辑是怎么工作的一种指导，只有人们知道了这一点，才能运用它用于做一些对自己有利的事情。
这本书就是一个指导，因此尽管你不是电气领域的专家，你也可以运用模糊逻辑。
需要指出的是有一些针对模糊逻辑的相反观点和批评。
一个人应该学会观察反面的各个观点，从而得出自己的观点。
我个人认为，身为被表扬以及因写关于模糊逻辑论文而受到赞赏的作者，他会认为，在这个领域中的这种批评有点过激。
编程语言外文翻译
编程语言编写计算机程序有两种常用方法：面向过程的程序设计和面向对象的程序设计。
面向对象的程序设计是面向过程程序设计的一种扩展，在编写程序时采用的方法有一些不同，用面向对象的方法考虑问题，首先把程序元素看成是与现实世界中的具体对象相似的对象，然后对这些对象进行操作以得到期望的结果。
编写面向对象的程序包括创建对象和创建使用这些对象的应用程序。
机器语言能被计算机操作系统直接运行的计算机程序称为可执行程序。
可执行程序是以机器码的形势表示的一系列非常简单的指令。
这些指令对于不同计算机的CPU 而言是特定的，它们与硬件有关。
例如，英特尔“奔腾”处理器和Power PC微处理器芯片各自有不同的机器语言，要求用不同的代码集来完成相同的任务。
机器码指令是从存储单元取数据，或将两个存储单元的内容相加（通常在CPU的寄存器中进行）。
机器码指令是二进制的——比特序列（0和1）。
由于这些数字令人难以理解，所以计算机指令通常不是用机器码来写的。
汇编语言与机器语言指令相比，汇编语言使用的命令较容易为程序员理解。
每条机器语言指令在汇编语言中有等价的命令。
例如，在汇编语言中，语句“MOV A,B”命令表示计算机把数据从一个单元复制到另一个单元，而及其代码中同样的指令是有一串16位的0和1组成的。
一旦汇编语言程序编写完毕，它就由另一个称之为汇编起的程序转换成机器语言程序。
相对于机器语言而言，汇编语言速度快，功能强。
可它仍然难以利用，因为汇编语言指令是有一系列抽象代码组成的。
另外，不同的CPU使用不同的机器语言，因此需要不同的汇编语言（程序）。
有时为了执行特殊的硬件任务，或者为了加快高级语言程序的速度，汇编语言被插入到高级语言程序中。
高级语言从机器语言进步到汇编语言，是语言达到了更先进的阶段。
同样也正是这种进步导致了高级语言的发展。
如果计算机能把简便的符号翻译成基本操作，为什么它就不能完成其他文字类型的编码功能呢？现在让我们来看看所期望的高级语言应有的特点，以及怎样将它们与机器码和汇编语言进行比较。
基于HMM的短语翻译对抽取方法
基于HMM 的短语翻译对抽取方法∗左云存 宗成庆中国科学院自动化研究所模式识别国家重点实验室 北京 100080E-mail: {yczuo, cqzong}@摘 要：在基于语料库的统计翻译方法中,基于短语的统计翻译与基于单个词的统计翻译相比可以更好地处理句中词语之间的关系，从而有效地提高机器翻译系统的性能。
在基于短语的统计翻译方法中,一种重要的策略是把短语翻译对作为一种知识加入到翻译系统中，因此,整个系统的性能与使用的短语翻译对的质量具有很大的关系。
本文在基于HMM 词对齐方法的基础上，提出了一种从双语语料中自动抽取短语翻译对的方法，这种方法根据词语对齐时出现的不同情况作不同的处理，提高了短语翻译对抽取的效果。
关键词：HMM ；词对齐；短语翻译对；机器翻译Phrase Translation Extraction Based on HMMZuo Yuncun, Zong ChengqingNational Laboratory of Pattern Recognition, Institute of Automation,Chinese Academy of Sciences, Beijing 100080E-mail: {yczuo, cqzong}@Abstract: In corpus-based statistical machine translation methods, phrase-based models are effective in improving translation quality as they can deal with the relationship between words in sentences better than word-based models. One approach of phrase-based translation incorporates phrase translations as knowledge sources into systems, and the systems’ performance greatly depends on the quality of phrase knowledge. In this paper, we describe a new approach of phrase translation extraction based on HMM-based word alignment method. The experiment result proved that this approach is effective in phrase translation extraction from bilingual corpus.Keywords: HMM, word alignment, phrase translation extraction, machine translation1 前言机器翻译的任务是把源语言句子(1...I s s s = )翻译成目标语言句子()。
translate
Lecture 1讲座一：HOW AND WHEN TO WRITE怎样写和何时写Scientific writing is part of the research work.科技写作时研究工作的一部分。
Research is not completed until it is published.直到写作完成，研究才算完整。
Scientific writing demands exactly the same qualities of thought that are needed in the rest of science: logic clarity and precision.科技写作需要同样准确的想法，这些想法是剩下的科学中需要的，清晰精确的逻辑。
Apply the scientific principals to the writing应用这写主要的科学的方法来写作。
Planing, Design, Execution计划，设计，实施Logically structured逻辑结构Clearly expressed清晰的表述Precisely worded准确的用词The way you write must follow from the way you work, not vice versa.你所写的必须与你的研究相对应，而不是反之。
In your scientific work you should have在你的科学工作中你应该有An observation观察A question followed by an hypothesis假设并提出问题How are you going to test the hypothesis? (Better adapt the negative approach).你怎样验证你的假设（最好适应相反的方法）You can't PROVE it. What you want to do is find data that will support or object your working hypothesis.你不能证明它。
体育专业课程思政建设应解决的问题及实施路径
第28卷第2期 2021年3月体 育 学 刊Journal of Physical EducationVol.28 No.2M a r.2021·学校体育·体育专业课程思政建设应解决的问题及实施路径董翠香1，樊三明2，李梦欣1，朱春山3，季浏1(1.华东师范大学体育与健康学院暨上海高校“立德树人”人文社会科学重点研究基地体育教育教学研究基地，上海200241；2.上海交通大学附属中学，上海200439；3.河南师范大学体育学院，河南新乡453007)摘要：运用文献资料、专家访谈、逻辑分析、实地调研等方法，剖释体育专业课程思政建设面临的困惑与应解决的问题，并以“学校体育学”为例，提出具体实施路径。
研究认为：体育专业课程思政建设应着重解决课程教学中重专业素养培育，轻人格修养培育；重知识技能传授，轻价值观培育；重低阶认知培养，轻高阶认知培养的“三重三轻”问题。
在此基础上，深入挖掘课程思政元素，并沿着课程目标引领、教学目标衔接、教学过程贯穿、课程评价落实的路径将课程思政元素落实到课程实施中，以提高体育专业课程思政教学效果。
关键词：学校体育；课程思政；课程思政元素；体育专业；学校体育学中图分类号：G807 文献标志码：A 文章编号：1006-7116(2021)02-0065-07The curriculum ideological and political construction of sports major shouldaddress the problems and the implementation pathDONG Cui-xiang1，FAN San-ming2，LI Meng-xin1，ZHU Chun-shan3，JI liu1(1.Research Base of Physical Education and Teaching in Shanghai “Lide Shuren” Key Research Base ofHumanity and Social Science，School of Physical Education and Health，East China Normal University，Shanghai 200241，China；2.High School Affiliated to Shanghai Jiao Tong University，Shanghai 200439，China；3.School of Physical Education, Henan Normal University, Xinxiang 453007，China)Abstract: By using the methods of literature, expert interview, logical analysis and field investigation, this paperexplains the puzzles and problems that should be solved in the curriculum ideological and political construction ofsports major, and takes “Physical Education Theory of School” as an example to put forward the specificimplementation path.The research shows that the curriculum ideological and political construction of sports majorshould focus on solving the following problems in course teaching as such: overemphasizing the cultivation ofprofessional accomplishment while underemphasizing the personality cultivation; overemphasizing the teaching ofknowledge and skills while ignoring the cultivation of values; overemphasizing the cultivation of low-levelcognition while underemphasizing the cultivation of high-level cognition. On this basis, the curriculum ideologicaland political elements should be deeply excavated, and following the path of curriculum goal guidance, teachinggoal convergence, teaching process from beginning to end, curriculum evaluation implementation to implement thecurriculum ideological and political elements into the curriculum implementation, in order to improve thecurriculum ideological and political teaching effect of sports major.Key words:school physical education；the curriculum ideology and politics；the curriculum ideological andpolitical elements；sports major；physical education theory of school收稿日期：2020-12-26基金项目：华东师范大学课程思政研究中心课题“体育教育专业课程思政教学指南研制”(2020KCSZ09)。
Kgdtys中学单词
-+懒惰是很奇怪的东西，它使你以为那是安逸，是休息，是福气；但实际上它所给你的是无聊，是倦怠，是消沉;它剥夺你对前途的希望，割断你和别人之间的友情，使你心胸日渐狭窄，对人生也越来越怀疑。
—罗兰a able about abovle accept across actor actress actual add adjective adverb advice advise affair after afternoon again against ago agree agreement ahead alive all almost alone along alond already also although altogether always among and anger angry animal announce another answer ant any anybody anyone anything anywhere apple April area army around arrive art artist as ask asleep at attention August aunt autumn awake away baby back bag ball banana bank base basic basin basket basketball bath bathroom be beach beat beautiful beauty because become bed bee beef beer before beg begin beginning behind believe bell below bench bend beside best better between bicycle big bike bill bird birth birthday biscuit bit bite bitter black blackboard blame blind blood blouse blow boat boil bomb bone born borronw both bottle bottom bowl boy brain bread break breakfast breath breathe bright bring broad broadcast brother brown brush build building burn bury bus bush business busy but butter buy by cab café cage cake call camera camp candle cap captain car card care careful careless carry case castle cat cause cave ceiling cent central centre century certain certainly china chairman chalk chance chat check cheek cheer cheese chemistry cheque chess chicken child chimney chocolate choice choose Christmas church cinema circle citizen city classmate classroom clean clear clever climb clinic clock close cloth clothes clothing cloud cloudy club coal coast cock coffee cold collar college color come comfort comfortable common communism communist complete computer consider continue control cook cool copy corn corner cost cottage cotton cough could count country countryside couple courage cousin cover cow crazy cream cross cruel cry cup cut damage dance danger dangerous dare dark darling date daughter dawn day dead dear death December decide decision deep deer defence defend delay delight depend depth desk destroy develop development dial dialond diamond diary dictionary die diet difference different difficult difficulty dig dinner direct direction director dirty disappoint discover discovery discuss discussion dish disk(disc) dislike distance distant district do doctor dog doll dollar door double doubt down downstairs dozen draw drawing dream dress drink driver drop drown deug dry duck during dust each eager ear early earn earth east eastern easy eat adit education effect egg eight eighteen eighth eighty either elder elephant eleven else email empty end enemy energy engineer enjoy enough enter entrance envelope equal equipment escape especially ever every everybody everyday everyone everything everywhere exam examination examine example excellent except exciting excuse exercise exhibition exist expet expensive experience experiment explain explanation eye face fact factory fall false family famoue far farm farmer farther fast fat father fault fear February feed feel feeling fellow fence fetch fever few field fifteen fifth fifty fight fill final find finger finish first fish fist fit five fix flag flight floor flower follow fond food fool foolish foot football for foreign foreigner forest forever forget forgive fork form fortunate forty forward found four fourteen fox free freedom freeze fresh Friday fridge friend friendly friendship frighten from front fruit full fun funny further future gain game garden gas gate gather generation gentle gentleman geography get girl give glad glass go goat God gold good goodbye goose government grade grandfather grandmother grass gray great green greet greeting grey ground group grow guard guess guest guide gun habit hair half hall ham handkerchief hang happy hard hardly harvest hat ahte have he headache headmaster health healthy heart heat heaven height hello help her here hers herself hi hide high hill him himself his history hit hold hole holiday home homework honest honour hope horse hospital host hotel hour housewife housework how however huge human hundred hunger hungry hurry hurt husband I ice-cream idea if ill illness imagine importance important impossible in inch income increase indeed information ink inside interesting international Internet into introduce invite island it its itself jacket January job join joke joy judge juiceJuly jump June just keep key kick kilogram kilometer kind king kiss kitchen kite knee knife knock know lab labour lack ladder lady lake land language large late lately later laugh law lawyer lay lazy leaf league learn least leave lecture left leg lend length less lesson let level lie life lifetime light like limit line lion lip list listen little lively living load lock look lose loss loud love lovely low luck lunch machine magazine mail mailbox main make man many map march maths matter May maybe me meal meat medical medicine meet meeting member memory message metal meter method metre midday middle midnight mile milk million mind miss mistake mister mix mode model modern moment Monday money month monthly more morning most mostly mother mountain mouse mouth move movie much music musical must my myself nail name narrow nation national native natural native natural near nearly neat necessary neck neeed needle neighhour neither nervous net never new news newspapper nice night nine nineteen ninety ninth no noble noon nor north northeast northern northwest nose not note nothing notice noun November now nowadays number nut ocean o’clock October of off office officer often oil okay old on once one oneself only onto open operate operation opinion opposite or oranger ordinary other otherwise ought our ours ourselves out outdoors over overcoat own ox page pain paint painting pair palace pale pan pardon parent particular partly party pass passage passenger past path pause pay peace peaceful peach pear peasant pen pencil penny people percent perfect perhaps person personal pet phone photo photograph phrase physics piano pick picnic picture pie pig pile pink pity place plan playground pleasant pleasure plenty plough pocket point police policeman polite political poor pop popular population pork port positoin possible possibly postcard poster postman pot potato practical practice practise praise prefer prepare price prince princess print prison prisoner probably problem produce production professor progress promiss pronunciation proper protect proud pud public pull punish pure push put quality quantity quarrel quarter queen question quick quiet quite rabbit radio rain raincoat rainy raise rather ray reach read reader reading ready real realize really recent record red reduce refuse regret regular relation remember remove repair repeat reply report republic research restaurant return rice rich ride right ring rise river road robot roll roof root rose round rubber rubbish rule ruler run sad safety sail sale salt same sand satisfy Saturday saw say scene science scientist sea search seat second secret secretary see seed seek seem seize seldom self sell send sense separate September serious servant service set seven seventeen seventy several sex shade shadow shake shall shame shape she sheep sheet shelf shine ship shirt shock shoe shop shore short shot should shout show shower shut shy sick side sight sign silent silk silly silver simple simply since sing sink sir sister sit situation six sixteen sixty size skate ski skill skin sky slave sleep slow small smart smell smile smoke snake snow so soap soccer sock sofa soft soldier solid some somebody someone something sometime sometime sometimes somewhere son song soon sorry soul soup south southeast southern southwest space speak speaker special speech speed spell spend spirit spoon sport sportsman spread square stair stand standard star start sate station stay steak steal steam steel step stick still stone stop store storey storm straight strange stranger street strength strict strike strong struggle study stupid style success successful such sudden sugar suitable suitcase summer sun Sunday super supermarket supper sure surface surprise sweater sweep sweet swim tail tailor take tale talk tall tape task taste taxi tea teach teacher team technical telegram telephone television temperature ten tennis tent term test text textbook than thank that the theatre their theirs them themselves then there these they thief thin thing think third thirst this those though thought thousand three through throughout throw Thursday thus tie tiger tight till time tin tiny tire to tobacco today toe together toilet tomato tomorrow ton tongue tonight too tool tooth top total touch tourist toward(s) town toy track trade traffic train training translate travel tree trick trip trouble trousers truck true trust truth Tuesday turn twelfth twelve twentieth twenty twice twin two type uncle uncle under understand university until up upon upstairs upward us use used useful useless user usual usually valley valuable verb very victory view village visit visitor voice volleyball wait waiter wake walk wall war warm warn wash waste watch wave way we weak weakness wear weather wedding Wednesday week weekdayweekend weigh weight welcome west western wet what whatever wheat wheel when whenever where wherever whether which while white who whole whom whose why wide wife wild will willing win window wine wing winter wipe wire wise wish with within wichout woman wonder wonderful wood wooden word work worker worry worse worst worth write wrong yard year yellow yes yesterday yet you young your yours yourself zero zoo。
abstract分析
Analysis of abstractAn abstract is a complete but concise statement that describes a larger work. Components of abstract are different from the discipline. The two most important reasons of abstracts are selection and indexing. Selection means that abstracts allow readers who may be interested in the long work to quickly decide whether this paper is worth their time to read it. Also, indexing means that many online databases use abstracts to index larger works. Therefore, the abstract is gaining tractions among us.There are two types of abstracts: Descriptive and Informative. A descriptive abstract outlines the topics covered in the paper. It makes no judgments about the work, nor does it provide results or conclusions of the research. The majority of abstracts are informative. An informative abstract includes the information that can be found in a descriptive abstract (purpose, methods, scope) but also includes the results and conclusions of the research and the recommendations of the author.From what I have mentioned above, the examples I choose (list of examples attached) can be divided into two: example 1 is a descriptive abstract, while the rests are all informative abstracts.The structure of a descriptive abstract may only include two parts like the example1: background and outline of the work. Usually, outline of the work likes a table of contents in paragraph form. Although descriptive abstracts are usually very short –100 words or less, it’s enough to know the structure of the whole work. Just li ke a miniature of the paper, it’s brief and complete.Example2~6 show the phenomenon that most abstracts are informative. An informative abstract often includes four parts: background, research question, method and conclusion. The background of an informative abstract is to answer the question that why do you want to do the research. Like example 2‖ Many practices aimed at cultivating multicultural competence in educational and organizational settings assume that multicultural experience fosters creativity‖, Is the assumption true? Or when and how does multicultural experience enhances creativity? With background of the topic, we will feel more comfortable to accept the conception of the work. The research question is the question the author trying to answer and the topic of the work. What problem you attempt to solve will determine whether this paper is worth the readers’ time.Like example 3‖This article investigates the possessions and activities that consumers love and their role in the construction of a coherent identity narrative‖, the question that example3 want to answer is the roles of loved objects and activities in the construction of a coherent identity narrative. The method of an informativeabstract is to answer the question that how do you get answers to your research question? While look at the list of examples, we’ll find that not all abstract show the approaches used in their work. The conclusion of an abstract is the work’s finding and value. Like example5 ―In an experiment, we show that vis ual personification—pictures in an ad that metaphorically represent a product as engaged in some kind of human behavior—can trigger anthropomorphism. Such personification, when embedded in an ad, appears to lead to more positive emotions, more positive attributions of brand personality, and greater brand liking. Implications for advertisers are discussed‖. From these words, we clearly know the final conclusion that visual personification can trigger anthropomorphism and personification in ads lead to positive effects.From the list of examples, we conclude that the language of abstract is brief and concise. Due to the limit of space, the abstract must be written the essential information in as simple terms as possible. Also, the work should be conveyed in as few words as possible with an assumption that it covers all the main point of the paper, that’s to say, the abstract should be complete. What’s more, the abstract ought to be reader-oriented, that’s means it should be written to serve reader in saving the ir time and energy.Here are some examples from the papers I have read recently:Example 1: Susan Fournier , Claudio Alvarez, Brands as relationship partners: Warmth, competence, and in-between, Journal of Consumer Psychology, October 2011.Background：The dialogue between social perception and consumer–brand relationship theories opens new opportunities for studying brands.Outline of the work:To advance branding research in the spirit of interdisciplinary inquiry, we propose to (1) investigate the process of anthropomorphism through which brands are imbued with intentional agency; (2) integrate the role of consumers not only as perceivers but also as relationship agents;(3) consider important defining dimensions of consumer–brand relationships beyond warmth and competence, including power and excitement; and (4) articulate the dynamics governing warmth (intentions) and competency (ability) judgments to yield prescriptive guidance for developing popular and admired brands.Example 2:Angela Ka-yee Leung, William W. Maddux ,Adam D. Galinsky, Chi-yue Chiu. Multicultural Experience Enhances Creativity:The When and How, American Psychologist, April 2008.Background：Many practices aimed at cultivating multicultural competence in educational and organizational settings (e.g., exchange programs, diversity education in college, diversity management at work) assume that multicultural experience fosters creativity.Research Question：In line with this assumption, the research reported in this article is the first to empirically demonstrate that exposure to multiple cultures in and of itself can enhance creativity.Conclusions:Overall, the authors found that extensiveness of multicultural experiences was positively related to both creative performance (insight learning, remote association, and idea generation) and creativity-supporting cognitive processes(retrieval of unconventional knowledge, recruitment of ideas from unfamiliar cultures for creative idea expansion). Furthermore, their studies showed that the serendipitous creative benefits resulting from multicultural experiences may depend on the extent to which individuals open themselves to foreign cultures, and that creativity is facilitated in contexts that deemphasize the need for firm answers or existential concerns.The authors discuss the implications of their findings for promoting creativity in increasingly global learning and work environments. Example 3: Aaron C. Ahuvia, Beyond the Extended Self: Loved Objects and Consumers’ Identity Narratives, Journal of Consumer Research, June 2005.Research Question：This article investigates the possessions and activities that consumers love and their role in the construction of a coherent identity narrative.Method：interviewConclusions: In the face of social forces pushing toward identity fragmentation, interviews reveal three different strategies, labeled ―demarcating,‖ ―compromising,‖ and ―synthesizing‖ solutions, for creating a coherent self-narrative.Findings are compared to Belk’s ―Possessions and the Extended Self.‖ Most claims from Belk are supported, but the notion of a core versus extended self is critiqued as a potentially confusing metaphor.The roles of loved objects and activities in structuring social relationships and in consumer well-being are also explored.Example 4:Nicolas Kervyn , Susan T. Fiske, Chris Malone, Brands as intentional agents framework: How perceived intentions and ability can map brand perception, Journal of Consumer Psychology, September 2011.Background：Building on the Stereotype Content Model, this paper introduces and tests the Brands as Intentional Agents Framework.A growing body of researchsuggests that consumers have relationships with brands that resemble relations between people.……Brands as Intentional Agents Framework is based on a well-established social perception approach: the Stereotype Content Model.Research Question：We propose that consumers perceive brands in the same way they perceive people. This approach allows us to explore how social perception theories and processes can predict brand purchase interest and loyalty.Conclusions: Two studies support the Brands as Intentional Agents Framework prediction that consumers assess a brand's perceived intentions and ability and that these perceptions elicit distinct emotions and drive differential brand behaviors.The research shows that human social interaction relationships translate to consumer–brand interactions in ways that are useful to inform brand positioning and brand communications.Example 5: Marjorie Delbaere, Edward F. McQuarrie, and Barbara J. Phillips, Personification in Advertising:Using a Visual Metaphor to Trigger Anthropomorphism, Journal of Advertising, Spring 2011.Background：All forms of personification draw on anthropomorphism, the propensity to attribute human characteristics to objects.Method：experiment,Conclusions: In an experiment, we show that visual personification—pictures in an ad that metaphorically represent a product as engaged in some kind of human behavior—can trigger anthropomorphism. Such personification, when embedded in an ad, appears to lead to more positive emotions, more positive attributions of brand personality, and greater brand liking. Implications for advertisers are discussed. Example 6:Marina Puzakova, Hyokjin Kwak,Joseph F. Rocereto, When Humanizing Brands Goes Wrong: The Detrimental Effect of Brand Anthropomorphization Amid Product Wrongdoings, Journal of Marketing, May 2013.Background：The brand relationship literature shows that the humanizing of brands and products generates more favorable consumer attitudes and thus enhances brand performance.Research Question：However, the authors propose negative downstream consequences of brand humanization; that is, the anthropomorphization of a brand can negatively affect consumers' brand evaluations when the brand faces negative publicity caused by product wrongdoings.Conclusions: They find that consumers who believe in personality stability (i.e., entity theorists) view anthropomorphized brands that undergo negative publicity less favorably than nonanthropomorphized brands. In contrast, consumers who advocate personality malleability (i.e.,incremental theorists) are less likely to devalue an anthropomorphized brand from a single instance of negative publicity.Finally, the authors explore three firm response strategies (i.e., denial,apology, and compensation) that can affect the evaluations of anthropomorphized brands for consumers with different implicit theory perspectives. They find that entity theorists have more difficulty in combating the adverse effects of brand anthropomorphization than incremental theorists. Furthermore, they demonstrate that compensation (vs. denial or apology) is the only effective response among entity theorists.。
2020年高考英语作文高分素材训练：4.2 人与社会
主题二 人与社会话题四 人际交往熟记50个高频单词·Group 1·1. acquaintance n. 熟人, (与某人)认识2. anxiety n. 担忧, 焦虑anxious adj. 忧虑的, 焦急的3. apologize vi. 道歉, 谢罪apology n. 道歉;歉意4. appreciate v. 欣赏;感激5. argue vi. 争辩, 争论argument n. 争论, 辩论6. associate vt. 联想;联系n. 同伴, 伙伴7. barrier n. 屏障, 障碍8. beneficial adj. 有利的, 有帮助的, 有用的benefit n. 益处, 好处v. 受益;有益于9. bother vt. 打扰, 烦扰10. colleague n. 同事·Group 2·1. comfort n. 安慰;慰问comfortable adj. 舒服的;安逸的;舒服自在的2. communicate v. 交际;传达(感情, 信息等) communication n. 交际, 交往;通讯3. companion n. 同伴;同事4. complain v. 抱怨;投诉5. conflict n. 冲突6. congratulate vt. 祝贺congratulation n. 祝贺, 庆贺7. conscience n. 良心, 良知, 内疚8. consider vt. 考虑considerate adj. 体贴的consideration n. 考虑;关心9. consult v. 请教, 商量10. conversation n. 对话·Group 3·1. correspond vi. 一致;与……相当;(与人)通信, 有书信往来2. debate n. &v. 讨论, 辩论3. dilemma n. (进退两难的)窘境, 困境4. disagree vi. 意见不一致disagreement n. 意见不一致;相违;争论5. embarrass vt. 使窘迫, 使尴尬6. equal adj. 平等的;vt. 等于, 比得上equality n. 平等7. exchange n. &vt. 交换, 调换;交流8. excuse n. 借口, 辩解vt. 原谅, 宽恕9. express vt. 表达;表示n. 快车, 特快专递expression n. 表达;词句;表示;表情10. favour(美favor)n. 恩惠;好意;帮助·Group 4·1. forgive vt. 原谅, 宽恕2. get-together n. 聚会3. grateful adj. 感激的, 感谢的4. greet vt. 问候;向……致敬greeting n. 祝贺5. harmony n. 融洽, 和睦6. invitation n. 邀请, 请帖invite vt. 邀请, 招待7. matter n. 要紧事, 事情;问题vi. 要紧, 有重大关系8. mature adj. 成熟的9. merciful adj. 仁慈的;宽大的mercy n. 怜悯10. misunderstand vt. 误会;不理解·Group 5·1. modest adj. 谦虚的;谦逊的2. offer n. &vt. 提供;建议3. patience n. 容忍;耐心4. persuade vt. 说服, 劝说5. promise n. &v. 答应, 允诺6. quarrel vi. 争吵, 吵架7. relate vi. 有关;涉及relation n. 关系;亲属relationship n. 关系relative n. 亲属, 亲戚8. reliable adj. 可信赖的, 可依靠的9. sensitive adj. 敏感的, 容易生气的10. trust vt. 相信, 信任, 信赖快记20个高频短语·Group 1·1. be ready to help others 乐于助人2. be disappointed with sb. 对某人感到失望3. be confident of sth. /in sb. 对某事/某人有信心4. be satisfied/content with对……满意5. be popular with受……欢迎6. be to blame应受责备7. be worried about担心……8. be concerned about 关心, 挂念9. complain about 抱怨10. communicate with与……交流·Group 2·1. get along/on well with与……关系良好2. help(sb. )out帮助(某人)摆脱困境3. have words with. . . 和……发生口角4. have a lot in common有许多共同之处5. keep company with和……结交6. make friends with sb. 和某人交朋友7. share feelings and ideas with与……分享感受和观点8. show respect for尊重9. think of others为别人着想10. work together共同努力牢记20个写作佳句1. Ours was a relationship that didn’t need a lot of words.我们的友谊是一种不需要很多话语的关系。
基于曲轴瞬时加速度分析的发动机启动过程着火判定与应用
基于曲轴瞬时加速度分析的发动机启动过程着火判定与应用3杨福源 张京永 王晓光 周 明 欧阳明高( 清华大学 ,汽车安全与节能国家重点实验室 ,北京 100084)[ 摘要 ] 提出不同于失火判定的基于曲轴瞬时加速度分析的发动机启动过程着火判定方法 ,即通过两次启动过程的瞬时加速度对比寻找喷油启动过程的着火始点 。
理论分析和试验表明该方法不仅可以准确识别着火始点 ,而且识别精度也从循环级提高到角度级 。
试验在一台 6 缸电控共轨柴油机上进行 。
叙词 :电控发动机 ,启动过程 ,燃烧始点判定 ,曲轴瞬时加速度J udgeme nt and Applicatio n of Ignitio n Mo m e nt Duri n g Engi ne St art upPerio d Based o n Crankshaf t Transie nt Acceleratio n Anal ysisYang Fuyuan , Z hang Jingyong , Wang Xiaoguang , Z hou Ming & Ouyang MinggaoTsi n g h u a U ni versi t y , S t ate Key L aborat ory of A ut om o t i ve S af et y a n d Energy , Bei j i n g 100084Abstract This paper brings fo rward an igniti o n detecti o n met h o d based o n crankshaf t t r ansient accelera 2ti o n analysis 1 It co m pares t h e t r ansient accelerati o n of t w o start u p p r ocedures and t h en find o u t t h e igniti o n m o 2ment ( start of co mbusti o n ) during start up peri o d 1Crankshaf t dynamics analysis and real test show t hat t h is met ho d can identi f y t he igniti o n m o ment ,wit h t he accuracy imp roved f ro m cycle level to angle level 1 The test is carried o u t o n a 6 - cylinder co m m o n rail diesel engine 1K ey w ords :E lectron ically controlled engine , S tartup period , I gn i t i on mom ent , C ran k shaf t transient acceler 2at i on将可对柴油机启动过程的喷油量 、喷油正时 、喷射压力等参数进行优化 。
新世纪大学英语综合教程1 课后答案 第二版 秦秀白unit1课后练习参考答案(P23-32)
Unit OneCollege EnglishEnhance Your Language AwarenessWords in Action▆Working with Words and Expressions1.Listed in the boxes below are some of the words that you need to be very familiar with. Now work inpairs and make sentences with each of them to see if you have really mastered their usages. You are encouraged to consult a dictionary if you are still not quite sure about their meanings and usages.▆ Sample sentences for reference:1) Do you think the government will deliver on their election promises?2) I need to polish (up) my French now because I’m going to France next summer.3) Details of the competition are available from our head office.4) Many people choose to go by car rather than by bus, but I prefer the latter.5) Please file this letter (away), Miss Smith.6) Today his company continues to thrive in spite of severe competition.7) She undertook responsibility for the changes he made in the article.8) A career advisor can offer you sound and practical suggestions on how to make career decision.9) A nurse has many duties to fulfill in caring for the sick.10)He perceived a subtle (微妙的；细微的) change in her manner when he met her the second time.11)Over the years, I have accumulated hundreds of books.12)He has multiplied his fortune many times within a short period of two years.2. In the box below are some of the words you have learned in this unit. Complete the following sentenceswith them. Change the form where necessary.▆ Answers:1)compromise2)self-induced3)steered4)frame5)demonstrated6)employ7)promote8)impressed9)contribution10)d eliberately11)f inancial12)e conomic3.In the boxes below are some of the expressions you have learned in this unit. Do you understand theirmeanings? Do you know how to use them in the proper context? Now check for yourself by doing the blank-filling exercise. Change the form where necessary.▆ Answers:1)makes a point of2)refresh my memory3)lead to4)at hand5)working out6)under pressure7)Last but not least8)took, down9)In addition to10)w ere involved in11)i n other words12)p ointed out13)p ay off4.Fill in the blanks with words listed in the box below. Change the form where necessary. Then rearrangethe given letters to form two words related to college life and complete the final sentence with the two words.▆ Answers:1)scored2)scheduled3)assigned4)motivate5)crucial6)promote7)performed8)debate9)scanned10)devised11)advocated12)clarify13)priorities14)compromised15)context16)undertook17)academic excellence▆Increasing Your Word Power1.The same word can perform different roles and functions as different parts of speech. Listed in the boxbelow are some of these words. Complete the following sentences with the words in the box. Make sure that your sentences are grammatically correct.▆ Answers:1)principal / major2)top3)major4)top5)principal6)major7)schedule8)advocate / have advocated9)top10)a pproach11)b lame12)m ajor / principal13)a dvocate14)s chedule15)b lame16)a pproaching17)p ressure18)p ace19)p ressured20)p ace2. In this unit there are many words and phrases that are similar in meaning. Find out those pairs ofwords and phrases and put them down in the space provided.▆3.In this unit the verb set is collocated with priorities, timetables, and standards. Try to fill in as manynouns as possible that can go with set in the space provided. If you are not sure of some collocations, consult a dictionary. Then make sentences of your own using the collocations in the space provided. ▆Answers for reference:set priorities / timetables / standards / an exam / an alarm / an example / a price /a record / limits1)Who is going to set (the questions for) the exam this time?2)She forgot to set her alarm last night: that was why she overslept.3)Her heroic behavior sets a good example for us to follow.4)We had to compromise and finally the price was set at $1,000.5)We were all pleased to know that Liu Xiang had set a world record in the 110-meter hurdles.6)The government has set strict limits on pay increases.4.Did you notice the suffixes -ion, -ation, -ly in words such as education, expectation, deliberately inthis unit? The suffixes -ion, -ation can be put after many verbs to form nouns, and -ly can be added to many adjectives to form adverbs. Listed in the box below are some nouns and adverbs with suffixes -ion, -ation, -ly you have come across in this unit. Write down their corresponding verb and adjective forms in the space provided.▆Task 1: Complete each of the following sentences with a proper noun with suffixes -ion, -ation listed above. ▆Answers:1)edition2)revisions3)contribution4)anticipation5)frustration6)participation7)association8)expectationsTask 2: Correct the mistakes in the following paragraph, paying special attention to the use of adjectives and adverbs. Underline the mistakes and write down your corrections in the space below. If youthink there is no mistake, write “No Mistake”.▆Answers:1)No mistake2)especially3)necessary4)frequently5)No mistake6)easy7)No mistake8)individual9)many10)h ighly11)a pparent12)r emarkable13)p robably14)N o mistakeGRAMMAR IN CONTEXTTask 1: Decide which one is correct by crossing out the wrong one.▆ Answers for reference:1)People go to university to increase their knowledge.2)The family were at table when the telephone rang.3)There is a Jack Robinson waiting to see you in the hall.4)We are striving for a world with no poverty, no hurt and no hatred.5)The Olympic Games for 2016 will be held in Rio de Janeiro.Task 2: Insert an appropriate article in each blank, using Ø to stand for ZERO article:▆ Answers:1)the2)an3)Ø4) a5)the6)the7)the8) a9) a10)a11)t he12)t he13)Ø14)t he15)t he16)a17)Ø18)t he19)t he20)aClozeComplete the following passage with words chosen from this unit. The initial letter of each is given.▆ Answers:1)a cademic2)p riorities3)c onducted4)principles5)begin6)p riority7)c ompromised8)a ddition9)filing10)S peaking11)f ormula12)Participation / Participating13)b ased14)least15)way16)pressureTranslationTranslate the following sentences into English, using the words and expressions given in brackets.▆ Answers for reference:1)The judge asked the reporters not to disclose the name of the victim.2)The teacher took pains to make sure that we all understood what he said.3)Recently the school conducted a survey among those students who have attained academic excellence.4)He said he would accept the job, so we have asked him to confirm his acceptance by writing us a letter.5)George studies very hard. He wants to make the most of his chance to learn.6)We can’t go. To begin with, it’s too cold. Besides, we’r e busy.7)It’s about time that someone spoke up for these basic truths / facts.8)You should be working instead of lying there in bed at this time of the day.9)I’ll jot down some notes while he’s speaking.10)I can’t carry the suitcase on my own; it’s too heavy.Theme-Related Writing1. Write an essay in about 100–120 words based on your discussion in Optional Classroom Activities. You can take either of the following titles:1) I like to study on my own2) I like to study in a group2. If you didn’t do Task 3 in Optional Classroom Activities you can write an essay of around 120 words on the following topic.College Is Not an Ivory Tower▆Sample Essays for Task 1:I like to study on my ownMany of my classmates like to study together. However, I always like to study on my own, with good reasons. First, I feel that I can concentrate better when I study alone. When I have to work together with other people, I often get distracted and more often than not, I end up chatting with them without completing the task at hand. Second, when I study on my own, I can work at my own pace, choosing to spend more time on certain difficult sections and less time on less challenging parts. I can also check my class notes or reference books whenever I feel the need to. One final good advantage of studying alone is that unlike group work, self study is not restricted by time and space. In other words, if I feel like studying, I can do so at any time and anywhere. (146 words)I like to study in a groupAccording to a recent study, students who study together by discussing homework and problems usually score higher than those who work on their own. I am happy to hear that because I like to study in a group. I think we can reinforce our understanding of what we have learned in class by asking each other questions. Besides, it is always enjoyable working with someone else. A discussion on questions we don’t understand very well will help clear up any possible confusion. Last but not least, we should not ignore the importance of cooperation in our society. Working with others helps develop our team spirit, which has become essential in this modern world of ours. After all, two heads are always better than one.(125 words)▆Sample Essay for Task 2:College Is Not an Ivory TowerPeople often say that college is an ivory tower and life on campus is cut off from the harsh realities of the real world. College may have been an ivory tower before, but it is not so any more. We college students are under various kinds of pressure. First of all, most of us are burdened with a heavy tuition fee, which keeps on increasing every year. So some of us have to work over ten hours every week. Life for those who can get financial help from their parents is not easy either, as they very often have to follow their parents’ advice and take a major they don’t like. Finally those who are about to graduate suffer from a different kind of pressure: hunting for a job. It is no longer easy for a college student to get a decently paid job these days.(134 words)。
维生素A的自述：还记得小时候被胡萝卜支配的恐惧吗？
62Univ. Chem. 2018, 33 (7), 62−65收稿：2018-03-02；录用：2018-03-20；网络发表：2018-03-26*通讯作者，Email: xinfengq@; wood@基金资助：四川大学双创专题研究项目(SCUCXCY1725)•科普• doi: 10.3866/ 维生素A 的自述：还记得小时候被胡萝卜支配的恐惧吗？操小桐1，权新峰2,*，吴迪3,4,5,*1四川大学生命科学学院，成都 6100642四川大学匹兹堡学院，成都 6102073四川大学化学学院，成都 6100644四川大学创新创业工作领导小组办公室，成都 6100645四川大学教务处，成都 610064摘要：文章以第一人称介绍了维生素A 的基本信息、生理功能、影响因素及相关生活常识。
关键词：维生素A ；胡萝卜；生理功能；抗癌；科普中图分类号：G64；O6The Readme of Vitamin A: Did You Remember the Fear Dominated by Carrots When You Were Young?CAO Xiaotong 1, QUAN Xinfeng 2,*, WU Di 3,4,5,*1 College of Life Sciences, Sichuan University, Chengdu 610064, P. R. China.2 Sichuan University-Pittsburgh Institute, Sichuan University, Chengdu 610207, P. R. China.3 College of Chemistry, Sichuan University, Chengdu 610064, P. R. China.4 Innovation & Entrepreneurship Office, Sichuan University, Chengdu 610064, P. R. China.5 Department of Academic Affair, Sichuan University, Chengdu 610064, P. R. China.Abstract: In the first person, this paper introduced the basic information, physiological functions, as well as some common sense of Vitamin A.Key Words: Vitamin A; Carrot; Physiological function; Anti-cancer; Science popularization做汤做菜，无不适宜。
不可逆温敏变色油墨的显色动力学探究
第43卷第11期包装工程2022年6月PACKAGING ENGINEERING·46·不可逆温敏变色油墨的显色动力学探究俞胡斐，钱静（江南大学，江苏无锡214122）摘要：目的探究温敏变色油墨对时间、温度的响应情况，以及不同制备配方对其动力学参数的影响。
方法利用光生酸剂的分解反应使结晶隐性紫染料开环显色，以此作为变色体系置于丝网油墨中，制备出可随时间–温度累积进行不可逆变色的智能温敏标签。
利用分光密度仪测量标签b*值，以标签b*值变化作为测量指标，将其与变色时间进行拟合，探讨油墨组分对标签变色性能的影响。
并探究不同组配方下b*值对温度变化的响应，以及不同配方对反应活化能的影响。
结果随着时间的积累，油墨印刷标签b*值逐渐减少，标签颜色由浅黄色变成深蓝色；同时，当温度升高，标签的变色速率也随之加快，温度越高，标签所测量的b*值越低，呈现出的蓝色越深。
相同温度下标签|Δb*|值较高时，标签的活化能较低。
结论该标签对时间、温度的积累均有响应，制备的标签活化能在11.453～27.676 kJ/mol波动，具有应用在食品品质智能监控领域的潜力。
关键词：不可逆变色；活化能；丝网印刷中图分类号：TS871 文献标识码：A 文章编号：1001-3563(2022)11-0046-08DOI：10.19554/ki.1001-3563.2022.11.007. All Rights Reserved.Color Kinetics of an Irreversible Thermochromic InkYU Hu-fei, QIAN Jing(Jiangnan University, Jiangsu Wuxi 214122, China)ABSTRACT: The purpose of this paper is to explore the response of thermochromic ink to time and temperature, and theeffects of different preparation formulas on its kinetic parameters. An intelligent thermochromic label that can discolorirreversibly with time-temperature accumulation was prepared by using the decomposition reaction of photo-acid genera-tor to discolor leucocrystal violet as a color-change system in screen ink. A spectrodensitometer was used to measure thelabel b* value, and make the change of label b* value as the measurement index. Then it was fitted with the discolorationtime to explore the influence of ink components on the discoloration performance of labels. With the accumulation oftime, the label b* value gradually reduced, the label color from light yellow to dark blue; at the same time, when thetemperature increases, the rate of discoloration of the label is accelerated. The higher the temperature, the lower the la-bel b* value measured, and the deeper the blue color presented. And the activation energy of the label is lower when thelabel |Δb*| value is higher at the same temperature. In conclusion, the label responds to the accumulation of time andtemperature, and the activation energy of the prepared label fluctuates from 11.453 to 27.676 kJ/mol, which has the po-tential to be applied in the field of intelligent monitoring of food quality.KEY WORDS: irreversible discoloration; activation energy; screen printing收稿日期：2021–08–03基金项目：国家重点研发计划重点专项（2018YFC1603300）；无锡科技局产业前瞻和关键技术（现代农业）（N20193008）；江苏省苏北专项（SZ–SQ2017049）作者简介：俞胡斐（1997—），女，江南大学硕士生，主攻智能包装。
我国体育教育训练学引文的学科结构状况及演变趋势
第28卷第3期 2021年5月体 育 学 刊Journal of Physical EducationVol.28 No.3M a y.2021·学校体育·我国体育教育训练学引文的学科结构状况及演变趋势王硕1，许治平2，郭伟1，张俊珍1，于善1，闵瑞欣1，张焱1，马琳1(1.西安体育学院，陕西西安710068；2.北京体育大学学术期刊社，北京100084)摘要：依据文献计量学相关理论和学科专业理论，从知识利用的视角对2015—2019年11种CSSCI体育核心期刊关于体育教育训练学引文的学科结构及演变趋势、亲缘关系及知识结构与模式进行深入剖析，以期探寻我国体育教育训练学科结构的构建特征及未来创新融合发展趋势。
研究表明：体育教育训练学是一门交叉融合创新发展较强的学科，知识建构的数据来源期刊间发文量和期刊自身年发文量均存在较大差异，知识建构的来源学科维度较大，且各学科维度对体育教育训练学科的影响差异也较大，但对其产生较大影响的学科却十分集中。
当前我国体育教育训练学科呈现出“双核蛛网式”4层知识结构模式，引文学科结构具有综合性但全面性不足、稳定性但独立性不足、固化性但深入性不足的特征，亲缘学科对体育教育训练学科产生重要影响的知识结构表现出偏于宏观性而微观性不足、维度较大而全面性不足的特征。
应不断加强和优化我国体育学科的知识建构，坚持走中国特色体育学科建设之路，逐渐演化成为一个以“体育教育训练学”单核中心向外延伸并辐射的多元、多“次中心”的综合性知识体系。
关键词：文献计量学；体育教育训练学；引文知识；学科结构；知识结构中图分类号：G80-05 文献标志码：A 文章编号：1006-7116(2021)03-0091-08Discipline structure condition and evolution trend in the citations forChinese physical education and trainingWANG Shuo1，XU Zhi-ping2，GUO Wei1，ZHANG Jun-zhen1，YU Shan1，MIN Rui-xin1，ZHANG Yan1，MA Lin1(1.Xian Physical Education University，Xian 710068，China；2.Journal Editorial Department，Beijing Sports University，Beijing 100084，China)Abstract:According to the relevant theories of bibliometrics and special discipline theory，and from the perspective of knowledge utilization，this paper makes an deep analyses on the subject structure, evolution trend, kinship and knowledge structure and mode of citations of sports education and training with 11 CSSCI sports core journals from 2015 to 2019, in order to explore the construction characteristics of the discipline structure of physical education and training and the trend of future innovation and integration development. The results show that physical education and training is a discipline with rather strong cross integration and innovation. There are great differences between the data sources of knowledge construction, the number of papers published between journals and the annual number of papers published by journals. The source discipline dimension of knowledge construction is relatively large, and the influence of each discipline dimension on physical education and training science may be also quite different, but the disciplines that has a great impact on it are very concentrated. At present, the discipline of physical education and training presents a "double core cobweb" four-layer knowledge structure mode. The收稿日期：2020-09-30基金项目：国家社会科学基金项目(17BTY023)；西安体育学院教学改革重点研究项目(XTJY1901)。
210975373_具身认知视阈下粤剧非遗数字化设计研究
第44卷 第6期 包 装 工 程2023年3月 PACKAGING ENGINEERING311收稿日期：2022–10–25基金项目：教育部产学合作协同育人项目（202102025002）；教育部产学合作协同育人项目（220900316224726）；教育部产学合作协同育人项目（202102199034）作者简介：汤晓颖（1977—），女，博士，教授，主要从事数字媒体传播与设计研究。
具身认知视阈下粤剧非遗数字化设计研究汤晓颖1，唐艺英2（1.广东工业大学，广州 510000；2.广东科技学院，广东 东莞 523000）摘要：目的 在具身认知理论的视阈下探讨粤剧非遗数字化设计的新模式，以改善因非遗文化与受众时空差距和传统观摩式展示造成的受众流失问题。
方法 主要采用分析法和用户旅程图法对现阶段粤剧非遗的传播现状及存在的痛点进行叙述，同时通过Citespace 软件分析出近几年具身认知运用到各个领域的关键词，利用具身认知理论提出复现、沉浸和编码的具身交互的设计模式，重构了非遗文化传播的知觉-身体-环境的关系，提供了一种沉浸感强、文化认同的设计方式。
结果用粤剧具身交互作品《穿我的新衣》和《变脸》实践了非遗文化数字化设计模式。
结论 粤剧文化数字化设计可以从具身认知角度切入，通过复现粤剧情景、沉浸式交互和文化认同式编码，给非遗文化数字化设计研究提供一种新的方法。
关键词：具身认知；粤剧非遗；数字化设计；具身交互中图分类号：TB472 文献标识码：A 文章编号：1001-3563(2023)06-0311-09 DOI ：10.19554/ki.1001-3563.2023.06.034Digital Design of Cantonese Opera from the Perspective of Embodied CognitionTANG Xiao-ying 1, TANG Yi-ying 2(1.Guangdong University of Technology, Guangzhou 510000, China;2.Guangdong University of Science and Technology, Guangdong Dongguan 523000, China)ABSTRACT: The work aims to discuss the new mode of digital design of Cantonese opera as the intangible cultural heritage based on the visual threshold of embodied cognition theory, in order to improve the audience loss caused by the time-space gap between intangible cultural heritage and audience and the traditional demonstration of observation culture. The present situation and pain points of Cantonese opera as the intangible cultural heritage in the current stage were mainly described by analysis and user journey diagram. At the same time, the key words of embodied cognition applied to various fields in recent years were analyzed by Citespace software, and the design mode of reappearance, immersion and coding embodied interaction was put forward by embodied cognition theory, which reconstructed the percep-tion-body-environment relationship of intangible cultural heritage communication, and provided a design mode with strong immersion and cultural identity. Finally, the digital design of intangible cultural heritage was verified by the inter-active works of Cantonese opera "Put on My New Clothes" and "Face Changing". The digital design of Cantonese opera culture can be cut from the perspective of embodied cognition to provide a new method for the research of digital design of intangible cultural heritage by reappearing Cantonese opera scenes, immersive interaction and cultural identity coding. KEY WORDS: embodied cognition; Cantonese opera as intangible cultural heritage; digital design; embodied interaction非物质文化遗产（非遗）是代表着人类精神文明高度的无形文化遗产，它的表现形式多样。
软交换系统的应用与分析
Telecom Power Technology通信网络技术软交换系统的应用与分析王昆，胡芳芳，傅星文92146部队，广东湛江524000在通信技术日新月异的时代，软交换系统因其稳定性强、灵活性高以及扩展性好等特点被广泛应用。
在需要依托软交换系统完成程控交换系统向IP多媒体系统从软交换的发展、结构、特点等方面进行介绍，与程控交换进行对比，基于具体设备进行组网方案设计。
程控交换；软交换；下一代网络（NGN）；SOFTX3000Application and Analysis of Softswitch SystemWANG Kun, HU Fangfang, FU Xingwen92146, Zhanjiang 524000Abstract: In the era of communication technology changing with each passing day, softswitch system is widely used because of its strong stability, high flexibility and good expansibility. In the field of automatic telephone network, itis necessary to rely on softswitch system to complete the smooth transition from program-controlled switching systemto IP Multimedia Subsystem (IMS). This paper introduces the development, structure and characteristics of softswitch, 2023年6月10日第40卷第11期· 167 ·Telecom Power TechnologyJun. 10, 2023, Vol.40 No.11王 昆，等：软交换系统的应用与分析中处理器的呼叫控制功能从交换机的软件中独立出来，将其升级为软交换功能。
结合视觉特征和场景语义的图像描述生成
第43卷第9期计算机学报Vol. 43 No. 9结合视觉特征和场景语义的图像描述生成李志欣1)魏海洋1)黄飞成1)张灿龙1)马慧芳1),2)史忠植3)1)(广西师范大学广西多源信息挖掘与安全重点实验室桂林 541004)2)(西北师范大学计算机科学与工程学院兰州 730070)3)(中国科学院计算技术研究所智能信息处理重点实验室北京 100190)摘 要现有的图像描述生成方法大多只使用图像的视觉信息来指导描述的生成，缺乏有效的场景语义信息的指导，而且目前的视觉注意机制也无法调整对图像注意的聚焦强度. 针对这些问题，本文首先提出了一种改进的视觉注意模型，引入聚焦强度系数自动调整注意强度. 在解码器的每个时间步，通过模型的上下文信息和图像信息计算注意机制的聚焦强度系数，并通过该系数自动调整注意机制的“软”、“硬”强度，从而提取到更准确的图像视觉信息. 此外，本文利用潜在狄利克雷分布模型与多层感知机提取出一系列与图像场景相关的主题词来表示图像场景语义信息，并将这些信息添加到语言生成模型中来指导单词的生成. 由于图像的场景主题信息是通过分析描述文本获得，包含描述的全局信息，所以模型可以生成一些适合图像场景的重要单词. 最后，本文利用注意机制来确定模型在解码的每一时刻所关注的图像视觉信息和场景语义信息，并将它们结合起来共同指导模型生成更加准确且符合场景主题的描述. 实验评估在MSCOCO和Flickr30k两个标准数据集上进行，实验结果表明本文方法能够生成更加准确的描述，并且在整体的评价指标上与基线方法相比有3%左右的性能提升.关键词图像描述生成；注意机制；场景语义；编码器-解码器框架；强化学习中图法分类号TP391 DOI号 10.11897/SP.J.1016.2020.01624Combine Visual Features and Scene Semantics for Image Captioning LI Zhi-Xin1) WEI Hai-Yang1) HUANG Fei-Cheng1) ZHANG Can-Long1)MA Hui-Fang1),2) SHI Zhong-Zhi3)1) (Guangxi Key Lab of Multi-source Information Mining & Security, Guangxi Normal University, Guilin 541004)2) (College of Computer Science and Engineering, Northwest Normal University, Lanzhou 730070)3)(Key Laboratory of Intelligent Information Processing, Institute of Computing Technology, Chinese Academy of Sciences, Beijing 100190)Abstract Most of the existing image captioning methods only use the visual information of the image toguide the generation of the captions, lacking the guidance of effective scene semantic information. Inaddition, the current visual attention mechanism cannot adjust the focus intensity on the image effectively.In order to solve these problems, this paper firstly proposes an improved visual attention model, whichintroduces a focus intensity coefficient so as to adjust attention intensity automatically. Specifically, thefocus intensity coefficient of the attention mechanism is a learnable scaling factor. It can be calculated bythe image information and the context information of the model at each time step of the language modeldecoding procedure. When using the attention mechanism to calculate the attention weight distribution on收稿日期：2019-08-27；在线发布日期：2020-02-07. 本课题得到国家自然科学基金(61966004, 61663004, 61866004, 61762078)、广西自然科学基金(2019GXNSFDA245018, 2018GXNSFDA281009, 2017GXNSFAA198365)、广西多源信息挖掘与安全重点实验室基金(16-A-03-02, MIMS18-08, MIMS 19-02)资助. 李志欣(通信作者)，博士，教授，博士生导师，中国计算机学会(CCF)高级会员，主要研究领域为图像理解、机器学习、跨媒体计算.E-mail:*************.cn.魏海洋，硕士研究生，主要研究领域为图像理解、机器学习. 黄飞成，硕士研究生，主要研究领域为图像理解、机器学习. 张灿龙，博士，教授，中国计算机学会(CCF)会员，主要研究领域为模式识别、目标跟踪. 马慧芳，博士，教授，中国计算机学会(CCF)会员，主要研究领域为机器学习、数据挖掘. 史忠植，研究员，博士生导师，中国计算机学会(CCF)会士，主要研究领域为人工智能、机器学习、神经计算、认知科学.9期李志欣等：结合视觉特征和场景语义的图像描述生成 1625the image, the “soft” or “hard” intensity of attention mechanism can be adjusted automatically byadaptively scaling the input value of softmax function through the focus intensity coefficient. Then theconcentration and dispersion of the visual attention can be achieved. Therefore, the proposed attentionmodel can make the extracted image visual information more accurate. Furthermore, we combineunsupervised and supervised learning methods to extract a series of topic words related to the image sceneto represent scene semantic information of the image, which is added to the language model to guide thegeneration of captions. We believe that each image contains several scene topic concepts, and each topicconcept can be represented by some topic words. Specifically, we use the latent Direchlet allocation (LDA)model to cluster all the caption texts in the dataset. Then the topic category of the caption text is used torepresent the scene category of corresponding image. What is more, we train a multi-layer perceptron (MLP)to classify the image into topic concepts. As a result, each topic category is represented by a series of topicwords obtained from clustering. Then the scene semantic information of each image can be represented bythese topic words, which are very relevant to the image scene. We add these topic words to the languagemodel so that it can obtain more prior knowledge. Since the topic information of the image scene isobtained through analyzing the captions, it contains some global information of the captions to be generated.Therefore, our model can predict some important words that suitable for image scene. Finally, we use theattention mechanism to determine the visual information of the image and the semantic information of thescene that the model pays attention to at each time step of the decoding procedure, and use the gatingmechanism to control the proportion of the input of these two information. Afterwards, both information iscombined to guide the model to generate more accurate and scene-specific captions. In the experimentalsection, we evaluate our model on two standard datasets, i.e. MSCOCO and Flickr30k. The experimentalresults show that our approach can generate more accurate captions than many state-of-the-art approaches.In addition, compared with the baseline approach, our approach achieves about 3% improvement on overallevaluation metrics.Keywords image captioning; attention mechanism; scene semantics; encoder-decoder framework;reinforcement learning1 引言图像描述是指根据给定图像的内容，为其生成合理的自然语言描述，是人工智能的一个重要研究领域，主要应用于图像和文本的相互检索、残障人士的生活辅助等方面. 图像描述包含对图像内容的识别以及自然语言生成方面的工作. 首先要求模型能够理解图像的内容，识别图像中的对象，并推理对象之间的关系等；其次是要求模型能够生成被人类理解的自然语言描述. 这是一个结合计算机视觉和自然语言处理的跨领域任务.通常图像中包含有大量显式和隐式的视觉语义信息，而图像和文本两种模态信息之间本身存在语义鸿沟，图像中的视觉信息实际上很难直接用自然语言完全表征. 最近大量关于图像描述的研究表明[1-3]，基于深度学习的方法可以很好地处理这一复杂任务. 这些方法通常基于一种来自机器翻译的编码器-解码器框架[4-6]，其主要思想是将图像描述任务看作是将一幅图像翻译成一段文本描述. 该框架一般使用卷积神经网络（Convolutional Neural Network，CNN）作为编码器进行图像编码，使用循环神经网络（Recurrent Neural Network，RNN）作为解码器生成文本描述. 这种方法在图像描述任务上取得了突破性的进展，因此目前编码器-解码器框架已经成为图像描述生成的基本方法.尽管取得了一些进展，但图像描述仍然是一项具有挑战性的任务，面临着若干需要重点考虑和解决的问题. 例如，如何更好地利用图像信息？如何更好地建立图像视觉特征与生成文本之间的联系？如何更准确地生成图像的描述语句？目前视觉注意机制已经被证明在图像描述任务中能够发挥很好的作用. 注意机制可以根据解码器的上下文信息，来1626 计算机学报 2020年重点关注在图像的一些显著区域，从而为词汇的生成提供精准有效的视觉信息指导. 目前的注意机制主要分为“软”注意和“硬”注意，在“硬”注意机制中模型只关注在图像最显著的一个区域上，去除了大量不必要信息的干扰，注意力权重分布是一个one-hot向量. 但由于其不能直接通过反向传播来进行训练，需要通过采样方法来估计梯度，相对来说比较复杂，因此目前大多数图像描述系统采用的是“软”注意机制[7-8]，即通过softmax函数来计算所有图像区域上的注意力权重分布，将所有图像区域的加权和作为视觉注意特征. 但由于softmax函数自身计算方式的缘故，如果传入softmax函数的数值区间较大，则生成的注意力分布就会相对集中，模型则变得比较“硬”，反之则较“软”. 而在生成描述的过程中，对于不同单词的生成，模型需要对图像施加不同聚焦强度的关注. 例如在生成一些名词时模型可能需要更“硬”的关注，集中地关注在图像的某个显著对象上；而在生成一些连词和介词时，可能需要较“软”的关注，分散地关注图像的所有区域. 此外，大多数图像描述系统缺乏场景语义信息的指导. 目前的图像描述系统对于图像的利用，大部分是直接通过CNN提取图像特征[1-3]或者是通过目标检测器提取图像上一些候选区域的特征[9-10]来表示图像信息，图像中潜在的场景语义信息却很少被利用. 而对于图像描述生成任务来说，场景语义信息对语句的生成至关重要. 因为对于同样的视觉特征，可能会有不同的场景含义，例如同样是草地，场景可能是公园的草地也可能是足球场的草地.针对以上问题，本文提出了结合视觉特征和场景语义的图像描述方法，图1是模型的整体结构. 首先，本文引入了聚焦强度系数来改进传统的视觉注意机制. 在每个时间步，通过模型的上下文信息和图像信息来计算视觉注意机制的聚焦强度系数，并通过该系数来自动地调控注意机制在每个时间步对图像区域的聚焦强度，从而使模型能够捕捉到更准确的图像视觉特征. 此外，本文将图像的场景语义信息添加到解码器中来指导单词的生成. 具体地说，对于图像中存在的场景主题概念，本文通过潜在狄利克雷分布（Latent Dirichlet Allocation，LDA）模型[11]来对数据集中的所有描述语句进行聚类分析，以描述语句的主题类别来表示图像的场景类别. 这里主题类别由聚类所得的一系列主题词汇来表示. 由于图像的场景主题信息是通过分析描述语句所得，因此可以获取到生成描述的一些全局信息，并且可以预知图像的描述语句可能包括哪些重要词汇，从而可对描述的生成提供很大的帮助. 在每个时间步中，通过注意机制来确定解码器重点关注的主题词，并结合图像的视觉特征，共同引导解码器生成更加准确和符合场景的描述. 最后，在标准数据集上对本文方法进行了测试，并在各项评估指标上与其他先进方法进行了对比. 结果表明，本文方法明显优于其他图像描述生成方法.图1 模型整体结构本文的主要贡献包括以下几个方面：(1) 提出了一种改进的视觉注意模型，通过自动调节注意机制的聚焦强度来提取更精准的图像视觉特征.(2) 通过将无监督和有监督方法相结合来进行图像场景语义信息的抽取，并使用注意机制将场景语义作为先验知识融合到解码器中.(3) 将图像视觉特征和场景语义相结合，来指导解码器生成更加准确和符合场景的描述.(4) 在MSCOCO和Flickr 30k两个标准数据集上进行了充分的实验验证，实验结果从定量和定性两方面证明了本文方法的有效性.9期李志欣等：结合视觉特征和场景语义的图像描述生成 16272相关工作目前大多数图像描述系统由图像编码器和语言解码器两部分组成. Mao等人[1]首次创造性地将机器翻译中的编码器-解码器框架应用于图像描述任务，并使用CNN作为图像编码器. 为了获得更好的解码能力，Vinyals等人[2]使用长短期记忆网络（Long Short-Term Memory，LSTM）替换普通RNN来为图像生成描述. 这种结构在图像描述任务上取得了突破性的进展，因此后续大量的研究人员一直基于此框架进行研究，并试图对编码器和解码器这两个部分进行改进.早期的图像描述编解码框架[1-3]只是简单地将CNN和RNN进行连接，将CNN最后的全连接层输出的向量作为图像的编码特征，在解码的初始时间步被输入到解码器中. 但这种方法只能得到图像的全局信息，不仅会丢失一部分图像特征，而且不能随着解码的进行有针对性地对图像进行解析. 为了能够更好地编码图像信息，Xu等人[7]在图像描述系统中引入了视觉注意机制，去除CNN最后的全连接层，将卷积层输出的特征向量作为图像的空间特征，并且通过“软”、“硬”两种注意机制来计算图像的空间注意力分布. 在每个时间步图像的注意特征与词嵌入向量拼接起来共同输入到解码器中来预测词汇的生成. 在生成不同词汇时，注意机制针对性地关注到图像的不同区域. 然而，在语句生成的过程中有些单词（例如：a, of）可能不需要关注图像的视觉信息，于是Lu等人[12]针对这个问题提出了一种基于视觉哨兵的自适应注意模型. 视觉哨兵可以确定何时关注图像信息，何时关注语言生成模型. 注意机制在图像描述任务中的运用，使解码器能够更好地整合图像信息，从而建立了更好的视觉信息与生成文本间的联系，使得图像描述任务取得了极大的改进. 目前大部分图像描述系统也都是基于注意机制和编解码框架相结合的结构.随着目标检测技术的发展，研究人员开始使用基于目标检测的编码器来提取图像特征，并且现在的语言解码器也变得越来越复杂. You等人[13]首先使用目标检测器从图像中提取一些视觉属性，然后将这些属性整合到语言模型中来增强视觉信息. Anderson等人[9]提出了一种自底向上和自顶向下的注意模型，使用Faster R-CNN[14]目标检测器来挑选一组具有高置信度的候选区域，并将这些区域的平均卷积特征作为图像的视觉特征. Lu等人[10]提出的神经婴儿谈话方法将早期的槽填充方法与基于神经网络的方法相结合，首先生成一个具有插槽的语句“模板”，然后通过目标检测器在图像区域中识别的视觉概念来填充这些槽. Gu等人[15]提出了由粗到细的多级预测框架，使用多个解码器来生成描述，每个解码器在前一级的输出上执行，产生越来越精细的描述语句. Jiang等人[16]提出了一种循环融合网络RFNet，利用多个编码器来提取图像特征，并通过多个LSTM进行信息间的循环融合，设计了非常复杂的信息交互.此外，强化学习和生成对抗网络（Generative Adversarial Net, GAN）[17]也逐渐应用到图像描述系统中，用以优化模型的生成结果. Ranzato等人[18]首先提出了一种基于RNN的策略梯度强化学习方法，直接在评价指标上优化模型的生成结果. Rennie等人[19]提出了一种自批评的强化学习方法，将模型在推理时生成的语句作为训练的基线，鼓励模型生成相对于基线更好的描述. Dai等人[20]则通过条件生成对抗网络[21]来生成多样化的图像描述. 强化学习主要针对模型的评价指标进行优化，使得模型的整体评估指标得分可以有较大的提高. 而基于GAN 的图像描述，大多关注的是描述语句的自然性和多样性，评价指标的得分可能反而较低.3模型本文采用统一的编码器-解码器框架来构建模型. 给定图像I，首先使用图像编码器来提取图像特征V，然后使用语言解码器进行逐步解码. 在每个时间步注意机制为解码器提供信息引导，最终解码器输出单词序列12{,,...,}TY y y y=（T是生成语句的最大长度）.3.1模型概述图像编码器：本文使用预训练的Faster R-CNN 从输入图像中提取一组候选区域特征作为图像的编码特征. 编码后的图像特征可以表示为=V12{,,...,}Lv v v，其中L是图像中候选区域的数量. 对于每个图像区域i，Ci∈v表示该区域的全局平均卷积特征. Faster R-CNN可以被认为是一种“硬”注意机制，它从整幅图像中挑选出相对少量的图像区域，可以去除一些不必要区域的干扰，与直接通过CNN提取特征相比，这种方法更具有针对性，具有明显的优势.语言解码器：在经典的编解码模型中，LSTM 通常作为解码器，用于构建语言生成模型. 在每个时间步t∈[1, T]，将图像的注意特征tV和上一时间步生成的单词1t y-，一同输入到LSTM中，输出1628 计 算 机 学 报 2020年LSTM 的隐状态t h ，然后通过t h 来预测单词的生成，LSTM 以此来逐步解码生成最终的描述序列. 1(,)t t Att -=V V h (1)1t e t y -=x W (2) 1([;],)t t t t LSTM -=h x V h (3) softmax()t t p t y =p W h (4)其中Att (⋅)表示视觉注意机制，用于计算图像的注意特征，t x 是在t 时间步输入的词嵌入向量，Ht ∈ h 是t 时间步LSTM 的隐状态，D Hp ⨯∈ W 用于将隐状态t h 映射到词典大小维度，Dt ∈p 是预测单词的概率向量（D 是包含所有单词的词典）.视觉注意机制：视觉注意机制源于对人类视觉的研究，其本质是对图像区域特征12{,,...,}L =V v v v 进行加权整合，即,1Lt t i i i α==∑V v ，在每个时间步将加权后的特征作为视觉信息输入到解码器中，在生成不同单词时，注意机制针对性地关注图像的不同区域. 权值分布是根据先前解码器的隐状态1t -h 和图像视觉特征V 来进行计算：,1tanh()T t i a v i h t e -=+W W v W h (5),,,1exp()exp()t i t i Lt i i e e α==∑ (6)其中K C v ⨯∈ W 和K H h ⨯∈ W 将图像特征i v 和隐状态1t -h 映射到同一维度，K a ∈ W . ,t i α是第i 个图像区域在t 时间步的注意权重. 3.2 改进的视觉注意模型在生成描述的过程中，由于不同词汇的特性不同，对于不同单词的生成，模型应该给予图像不同聚焦强度的关注来提取图像的视觉信息. 如式(6)所示，传统视觉注意机制通过softmax 函数来计算图像各个区域的注意分布，如果输入e 的数值区间较大，则经过指数化后，会进一步拉大数值间的差距，最终输出的权重分布就会相对集中；反之，权重分布则相对分散. 有鉴于此，本文提出了自适应聚焦强度的视觉注意机制，通过自动调整e 的数值区间来使模型能够自动调控视觉注意力的聚焦强度，从而能够针对不同的生成单词更好地提取图像的视觉信息.本文设置了一个聚焦强度系数η来控制视觉注意机制的聚焦强度. 在每个时间步随着上下文信息的变化，t t βηλ=可以自动的调整模型的聚焦强度，其中λ是设定的超参数，t β通过模型自身学习得到，具体地说是通过图像信息和模型的上下文信息来进行计算：1tanh(+)T T t b d t β-=W V W h (7)其中C b ∈ W ，H d ∈ W . V 表示图像的所有区域的平均特征，1t -h 是解码器上一时间步的隐状态. 将聚焦强度系数添加到式(6)中，可得,,,1exp()exp()t t i t i L t t i i e e ηαη==∑ (8) 改进后的视觉注意机制和编解码框架相结合可构建基于视觉注意的图像描述模型. 这里采用文献[9]提出的解码器结构作为语言解码器，其构造如图2所示. 与传统解码器不同的是，它由两个LSTM 组成，其中V-LSTM 为表示视觉注意LSTM ，用于整合当前信息，并为注意机制提供上下文信息输入；L-LSTM 表示语言LSTM ，用于预测单词生成，IV-Att 是改进后的视觉注意机制，具体的解码操作如下所示：11([;],)V V L Vt t t t LSTM --=h V x h h (9)(),V t u t Att =V W V h (10) 1([;],)L L V L t t t t LSTM -=h V h h (11)softmax()L t t p t y =p W h(12)图2 改进的视觉注意模型在每个时间步V-LSTM 接收图像的平均特征V和词嵌入向量t x ，以及模型的历史信息，并将这些信息进行整合输出隐状态Vt h . 然后，V t h 与图像特征V 一同输入到改进后的注意力模块IV-Att ，通过注意机制来计算图像的注意特征V t , 其中H C u ⨯∈W . 得到的视觉注意特征t V 与上下文信息V t h 一同被输入到L-LSTM 并输出Lt h ，最终通过Lt h 来预测当前时间步要生成的单词t y . 3.3 场景语义信息提取目前，大多数图像描述系统缺乏场景语义信息的指导. 而对于图像描述生成任务来说，场景语义信息对语句的生成至关重要. 本文的图像场景语义信息提取方法可分为三个步骤，如图3所示.9期 李志欣等：结合视觉特征和场景语义的图像描述生成 1629图3 场景语义信息提取方法首先，使用无监督的方法对数据集中所有的描述文本进行聚类. 具体的，将数据集中每张图像的多个描述语句合并为一个文档，然后使用LDA 模型来对所有的描述文档进行聚类分析. 数据集中的所有描述文档被划分为N 个主题类别，每个主题类别都通过一系列主题词汇来表示，这里选取具有最高概率的M 个词来表示一个主题类别，即：i U =,1,2,{,,...,},(1,)i i i M w w w i N ∈. 分类后，描述文档的主题类别可以看作其对应图像的场景主题类别，这样就可以得到了数据集中所有图像的场景类别标注.然后，利用标注后的图像数据，可通过监督学习来训练一个多层感知器（Multi-Layer Perceptron ，MLP ）. 输入图像的视觉特征，MLP 可输出图像对应的场景主题类别.第三步是通过训练好的多层感知器来对没有描述语句的图像进行场景主题分类. 分类后的每张图像都有一个对应的主题类别U i ，主题类别U i 通过一系列主题词w 来表示，这样就得到了图像的场景语义信息. 而在这些语义信息中包含了该图像对应描述的一些重要词汇，这些信息可以使解码器能够预知到一些需要生成的词汇，并对生成描述的全局信息有一定的掌控，对于提升模型性能会有很大的帮助.3.4 结合视觉特征和场景语义为了使模型能够生成更加准确且符合图像场景的描述，本文提出了结合图像视觉特征和场景语义的图像描述生成方法. 通过一系列与图像场景密切相关的主题词汇来表示图像的场景语义信息，并将其添加到语言模型中，与图像的视觉信息相结合，从而使模型能够得到更丰富的图像信息，并提前预知到一些需要生成的重要词汇，以此来共同引导模型生成更加准确且符合场景的描述.本文方法是基于图2的结构进行拓展，将图像的场景语义信息添加到解码器中. 图4是模型解码器的结构图，其中S-LSTM 表示场景LSTM ，S-Att 表示场景语义注意机制. 具体的解码操作如下：e i U =S W (13)11([;;],)S S L St t t t LSTM --=h S x h h (14)=(,)S t t Att S S h (15)其中U i 表示图像的场景主题类别，它由M 个主题词汇组成，首先将其转化成词嵌入向量形式∈SM H ⨯ ，S 表示M 个主题词汇的平均特征. 与V-LSTM 类似，S-LSTM 用于整合当前时刻输入的信息和历史信息，并为注意机制提供上下文信息输入S t h . 然后将S t h 和S 共同输入到S-Att 模块中，通过S-Att 输出当前时间步模型关注的场景语义信息.图4 结合视觉特征和场景语义的图像描述模型此外，这里设置了一个控制门，来控制当前信息的输入. 在提取到的场景语义信息中实际上包含了一些重要的视觉语义和描述中可能存在的一些重要词汇，这些词汇可以看作是视觉信息额外的补充，可以和视觉信息共同指导单词的生成. 与文献[12]的方法不同，本方法的控制门主要用来控制信息的输入量，以避免引入过多的重复信息或者其他的干扰信息. 具体操作如下：([;;;])V S t g t t t t g σ=+g W h h V S b (16),(1)t t t t t t ==-V g V S g S (17)1([;;;],)L L V S L t t t t t t LSTM -=h V S h h h (18)softmax()L t t p t y =p W h (19)其中H t ∈ g ， 表示对应位置元素相乘的操作. 本方法最后将图像的视觉注意信息t V ，场景语义注意信息t S ，隐状态V t h 和S t h 一同输入到L-LSTM 中，输出Lt h 用于预测当前时刻生成的单词.在本方法中，图像的场景语义信息作为先验知识添加到语言模型中，使模型可以获取更多的图像信息，并且具有更强的全局建模能力. 同时在每个时间步，场景语义注意机制可以使模型获取一些重要词汇信息，结合图像的视觉注意信息，以此来使模型生成更加准确且符合场景的描述.1630 计 算 机 学 报 2020年3.5 训练目标在模型的训练中，首先使用交叉熵损失来进行训练. 通过给定训练图像的正确参考语句的单词序列12{,,...,}T Y y y y =来最小化模型的交叉熵损失：1()log ()Tt t L p y θ==-∑ (20)即在每个时间步最大化正确参考单词的概率.然而基于交叉熵损失的训练存在严重的问题. 首先，模型在训练时每个时刻输入的是正确的参考单词，通过交叉熵损失来最大化下一个正确单词的概率，但在推理阶段模型依赖自身之前生成的单词来预测接下来单词，这样在训练和推理之间就存在一种“暴露偏差”[18]. 也就是说，在推理过程中一旦模型前面的单词生成的不好，就会导致误差累计，使得后面的单词也不会很好地生成. 此外，评估模型的评估指标也与交叉熵损失不相关，即模型训练和测试过程的目标不一致. 为了解决这些问题，可以将图像描述生成过程转化为一个强化学习问题，通过强化学习方法直接在评价指标上优化模型的生成效果. Rennie 等人[19]的研究表明，优化模型的评价指标CIDEr[22]可以使模型所有的评估指标得分都会有所提升. 因此，为了能够得到更好的生成结果，本文也使用了强化学习方法针对评价指标CIDEr 对模型进行了进一步的优化. 基于强化学习方法的训练目标是最大限度地减小负奖励期望：()[()]()Y p L E r Y r Y θθ=-≈- (21) 其中r (Y )表示模型生成语句Y 的CIDEr 得分. 梯度()L θθ∇可以通过蒙特卡罗方法近似估计：~()[()log ()]()log ()L E r Y p Y r Y p Y θθθθθθθ∇=-∇≈-∇Y p (22) 本文遵循Rennie 等人[19]提出的SCST 训练方法，使用模型在推理时生成的语句ˆY作为基线，来强迫模型生成相对于基线语句更好的描述. 即：ˆ()(()())log ()L r Y r Y p Y θθθθ∇≈--∇ (23) 基于强化学习的训练方法使得模型在训练和推理过程中保持一致，解决了图像描述模型中存在的“暴露偏差”[18]问题. 更重要的是，它直接在评价指标上优化了描述的生成，从而使模型在训练目标和测试指标上也保持一致，极大地提升了模型的整体性能.4 实 验为了证明本文提出方法的有效性，在MSCOCO 和Flickr30k 两个标准数据集上进行了充分的实验验证，将其与当前先进的模型进行了对比，并从定量和定性两个方面进行了结果分析. 4.1 数据集和评估指标本文在MSCOCO 和Flickr30k 数据集上进行实验来评估提出的模型，数据集划分方式如表1所示.由于MSCOCO 数据集的测试集中没有标注语句，因此采用Karpathy [3]对MSCOCO 数据集的划分方式，从验证集中挑选出5000张图像用于验证，5000张图像用于测试，验证集剩余的图像与训练集一起用作训练数据，数据集的每张图像包含5个人工标注的描述语句. 本文通过对数据集的图像描述文本分析来对文本进行预处理，用“UNK ”替换掉出现次数少于5次的低频单词.表1 数据集划分数据集训练集验证集 测试集 Flickr30k 28000 1000 1000 MSCOCO 11328750005000图5展示了不同语句长度在各自数据集中所占的比重，可以看出MSCOCO 数据集中的语句长度大多集中在8到15个单词之间，因此在MSCOCO图5 数据集中语句长度分布图。
系统框图化简的一种易忽略的思路
0引言近年来，在控制系统的分析和设计过程中，经常要计算一个复杂控制系统中某输入点到输出点的传递函数表达形式，并将它作为新的“单元”加入一个更大的系统中；或者对这个组合结构进行整体的考察和分析[1]。
另外，几乎每年的全国自动控制类课程自学考试中也都出现了框图化简并求传递函数的试题。
对于可化简的单输入-单输出线性系统框图，都能利用常规的移动引出点和比较点的方法进行等效化简，但采用何种移动思路才能顺利化简，至今还无相关论述，致使很多人尤其是初学者对复杂框图的化简只能乱打乱撞；对于由复杂框图求传递函数，常用的方法由框图化简法、梅森公式法、较死板的解析法等，其中的框图化简法较易掌握，且既一目了然又方便检查。
本文介绍了一种易忽略的框图化简思路，将复杂可化简组合框图进行等效变换，以方便计算系统的传递函数，进而分析系统的相应性能。
1框图的基本连接及等效变换系统框图常用来描述用传递函数方式表示的控制系统，是控制系统原理图与数学方程的结合，是线性定常系统常用的一种数学模型。
它表示一个动态“单元”把输入信号“放大”成为输出信号的过程，而“放大增益”即为该“单元”的传递函数[2]。
利用系统框图的串联、并联和反馈组合，多个框图可以组合表示一个复杂的物理系统，而方便且常用的方法是通过对框图的等效变换和化简来计算传递函数，可未形成一种行之有效的等效变换思路却不能如愿以偿。
控制系统的框图，由带箭头的信号线、引出点、比较点及功能框四种基本单元组成，可以形象地描述自动控制系统各单元之间和各作用量之间的相互联系，具有简明直观、运算和检查方便的优点[3]。
图1即为系统框图的三种基本连接方式：串联、并联和反馈，以及各基本连接的等效变换。
图1系统框图的三种基本连接及等效变换实际中，一个可以化简的系统框图，往往并不能直接使用如图1所示的基本等效变换来达到最简化目的，而是必须再移动引出点和比较点。
2框图等效变换的思路为实现等效变换和框图简化的顺利进行，移动引出点与比较点有如下一种易忽略的思路：2．1只将引出点移到与之最近但其间无比较点的功能框之前或之后，旨在使某两引出点能直接进行前后位置交换（如图2（a）所示），以使图1中至少两种基本连接能接二连三的出现，从而使否则移动的次数增加甚至无法达到最简化；图2相邻引出点及比较点的位置交换2.2只将比较点移到与之最近但其间无引出点的功能框之前或之后，旨在使某两比较点能直接进行前后位置交换（如图2（b）所示），以使图1中至少两种基本连接能接二连三的出现，否则将走向“简化”的反面；2．3尽量不要利用等效变换的原则将引出点与比较点在空间上交换前后位置，这样只会使框图的结构更复杂，而最后也只能用解析法求出其传递函数。
课标单词定
daily --adj. 每日的，日常的danger --n. 危险dangerous --adj.有危险的不安全的dark --adj. 昏暗的，黑暗的daughter --n. 女儿die--v. 灭忙，死忙，消失dead--adj. 死的，失去生命的deaf--adj. 聋的deal--v. 对付，对待decide--v. 决定，选定decision--n. 决定，抉择degree--n.（大学）学位,度数, 程度depend--v. 依靠，依赖describe--v. 描述develop--v. 发展，壮大development--n. 发展,发育, 成长dialogue--n. 对话，对白diary--n. 记事簿，日记dictionary--n. 词典，字典different--adj. 不同的difference--n. 差别，差异difficult--adj. 困难的difficulty--n. 困难，难题dig--v.挖（土）,掘（地）,凿（洞）direct--adj.直接的，直率direction--n. 方向director n.导演,部门负责人dirty--adj. 脏的discover--v. 发现，发觉discovery--n. 发现，发觉discuss--v. 讨论，商量discussion--n. 商量，讨论divide--v. 分开，分散double--v.加倍,是...的两倍adj.两倍的,加倍的doubt--v. 怀疑n.疑惑,疑问drop--v. 落下,掉下earthquake--n. 地震east--adv. 朝东，向东east adj. 东方的，东部的n. 东，东方eastern--adj. 东方的，东部的educate--v. 教育,教导education--n. 教育either--conj. 或者adv. 也（用在否定词组后）electronic--adj.电子的,电子设备的else--adv. 别的，其他的email--v. 用电子邮件发送，发电子邮件n. 电子邮件empty—adj.空的，空洞的encourage--v. 鼓励enemy--n. 敌人，仇人engineer--n. 工程师enter--v. 进来，进去environment--n. 环境eraser--n. 橡皮especially-adv.尤其,格外,特别Europe--n. 欧洲European--adj.欧洲的,欧洲人的n. 欧洲人exactly--adv. 确切的,精确的excellent--adj.极好的,优秀的except--prep. 除...之外conj. 除了，只是excuse--v. 原谅，宽恕exercise--v./n. 锻炼，练习expect--v. 预料，期待experience--n. 经验，经历express--v. 表达，表示explain--v. 解释，说明factory--n. 工厂fail-v.不及格,失败,未能（做到）fair--adj.合理的，公正的n. 展览会,交易会fantastic-adj.极好的,了不起的favorite--adj.特别喜爱的（人或事物）fear--v./n. 害怕，恐惧field--n. 田野，场地fight--v./n. 打架，战斗fill--v. (使)充满，装满finger--n. 手指fisherman--n.渔民,钓鱼的人fit--v. 适合，合身flag--n. 旗帜，旗force--n. 力，力量foreign--adj. 外国的form--n. 形式，类型France--n. 法国French--n. 法语fresh--adj. 新鲜的,清新的garden--n. 花园general--adj.普遍的,常规的,总的n. 将军gentleman--n. 先生，绅士geography--n. 地理，地理学German--n. 德国人，德语adj. 德国的，德国人的Germany--n. 德国giraffe--n. 长颈鹿glue--n. 胶水gold--adj. 金色的n. 金子,金币government--n. 政府，内阁grammar--n. 语法grape--n. 葡萄grey--adj. （天空）阴沉的，灰暗的，灰色的guard--n. 警卫，看守v. 守卫，保卫guest--n. 宾客，客人guide--n. 导游，向导guitar--n. 吉他gun--n. 枪，炮v差别habit (n.) 习惯half (n.& pron.) 一半；半数hall (n.) 大厅；礼堂hamburger (n.) 汉堡包handbag (n.) 小手提包handsome (adj.) 英俊的hang (v.) 悬挂；垂下happen (v.) 发生；出现hard (adv.) 努力地adj困难的hardly adv几乎不,几乎没有harmful (adj.) 有害的headache (n.) 头疼health (n.) 健康healthy (adj.) 健康的healthily (adv.) 健康地hear (v.) 听到，听见heavy (adj.) 重的heavily (adv.)在很大程度上；大量的high (adj./adv.) 高的（地）height (n.) 身高，高度helpful (adj.) 有用的；有帮助的heat (n.) 热，高温(v.)加热，变热hero (n.) 英雄，男主角hide (v.) 隐藏，隐蔽hill (n.) 小山，山丘hurt (v.)（使）疼痛，受伤husband (n.) 丈夫ice-cream (n.) 冰激凌ill (adj.) 有病，不舒服illness (n.) 疾病，病imagine (v.) 想象，设想important (adj.) 重要的importance (n.)重要性，重要impossible adj不可能发生的或存在的，不可能的include (v.) 包括，包含increase (v.) 增长，增加India (n.) 印度Indian (adj.) 印度的(n.) 印度人industry (n.) 工业，行业influence (v./n.) 影响information (n.)信息，消息inside (prep./adv.) 在里面instead (adv.)代替，反而，却instruction (n.)指示，命令instrument n. 器械,仪器,工具interest (n.) 兴趣，关注(v.)使感兴趣，使关注interested (adj.) 感兴趣的interesting (adj.)有趣的international (adj.)国际的Internet (n.)英特网，互联网introduce (v.) 介绍，引荐introduction (n.) 介绍interview (v.) 采访，面试(n.)面试，访谈invent (v.) 发明，创造invention (n.)发明，发明物invite (v.)邀请island (n.)岛January (n.) 一月Japan (n.) 日本Japanese (adj.) 日本的，日本人的，日语的(n.)日本人，日语join (v.)参加，加入journey (n.)（尤指长途)旅行；行程juice (n.)果汁；饮料June (n.)六月July (n.)七月just (adv.)只是，恰好keep (v.)保持，保留keeper (n.)饲养员，保管人key (n.)钥匙，关键keyboard (n.)键盘式电子乐器kick (v.)踢，踹kid (n.)小孩，年轻人(v.)开玩笑，欺骗kilo=kilogram (n.)千克，公斤kill (v.)杀死，弄死kind (adj.)和蔼的(n.)种类kilometer (n.)千米，公里king (n.)国王，君主kitchen (n.)厨房knee (n.)膝，膝盖knife (n.)（pl. knives）刀knock (v.)敲，击(n.)敲击声，敲击knowledge (n.)知识，学问laboratory (=lab.)实验室lake (n.)湖，湖泊life (n.)生活，生命lift (v.)举起，抬高(n.)电梯，搭便车light (n.)光，光线，光亮(v.)点亮，点燃(adj.)轻的，浅的like (v.)喜欢，喜爱(prep.)像，怎么样line (n.)行，排list (v.)列表，列清单(n.)名单，清单litter (n.)垃圾，废弃物(v.)乱扔live (v.)居住，生活(adj.)现场的；直播的lively (adj.)生气勃勃的，（色彩）鲜艳的local (adj.)当地的，本地的(n.)当地人，本地人lock (v.)锁上，锁住London(n.)伦敦lonely (adj.)孤独的；寂寞的lose (v.)失去，丢失，输lost（adj.)丢失的；迷路的loud (adj.)响亮的，大声的loudly (adv.)喧闹地，大声地lovely (adj.)可爱的low (adj.)低的，矮的luck(n.)幸运，运气lucky (adj.)幸运的luckily (adv.)幸运地，好运地lunch (n.)午餐machine (n.)机器,机械,装置mad (adj.)很生气的，疯的magazine (n.)杂志magic (adj.)有魔力的，有神奇力量的magician(n.)魔术师manage (v.)完成（困难的事）；manager (n.）经理应付（困难局面）March (n.)三月mark (n.)分数，记号，迹象(v.)做记号，打分market (n.)市场，集市marry (v.)结婚match (v.)匹配(n.)火柴matter (n.)问题，事情meal (n.)早（或午，晚）餐；一餐所吃的食物maybe (adv.)大概,或许,可能mean (v.)意思是,打算,意欲(adj.)吝啬的meaning (n.)意义，意思medical (adj.)医学的,医疗的medicine (n.)药，医学meet (v.)遇见，相逢meeting (n.)会议,集会,会面member (n.）成员，分子mention (v.)提到，说到mess (n.)杂乱，不整洁message (n.)信息，消息method (n.)方法，措施middle (n.)中间(adj.)中间的might (modal.v.)可能，可以mind (n.头脑，心智( v.)介意，注意minute (n.)分钟missing (adj.)丢失的mistake (n.)错误，失误(v..)弄错mix (n.) 混合配料(v.)（使）混合,融合model (n.)模型，模特modern (adj.)当代的,现代的month (n.)月份mouth (n.)嘴mountain (n.)高山mouse (n.)老鼠，耗子（pl. mice）move (v.)移动museum （n.)博物馆musician (n.)音乐家national adj.国家的;民族的natural adj. 自然的nature n. 自然界；大自然nearly adv. 几乎necessary adj.必要的;必须的neighbour/neighbor n.邻居neither adv.也不,两者都不去nervous adj.焦虑的;担忧的nineteen num. 十九ninety num.九十ninth num. 第九noisy adj. 吵闹的November n. 北部的notice n. 通知；通告；注意v.注意到；意识到nurse n.护士object n.物体；物品offer v. 主动提出;自愿给予officer n. 军官；官员Olympics n.奥林匹克运动会once adv. 一次；曾经online adv./adj. 在线(的);联网(的)opposite prep. 与…相对；与…对面adj. 对面的，另一边的order n./v.点菜；命令over prep.超过；多余；在......上方own v. 拥有；有Pacific n. 太平洋palace n. 宫殿；王宫pancake n. 烙饼；薄饼passage n. 章节；段落passenger n. 乘客；旅客passport n. 护照patient n. 病人adj. 有耐心的perfect adj.完美的；完全的perhaps adv.可能;大概;也许period n. 一段时间;时期personal adj.个人的;私人的physics n. 物理；物理学picnic n.野餐pilot n. 飞行员pioneer n. 先锋；先驱pity n.遗憾；怜悯v.同情；怜悯planet n. 行星playground n.运动场;操场please interj.(用于客气地请求或吩咐)请v.使高兴；使满意pleasure n. 高兴；愉快plenty pron. 大量；众多point n. 得分；点police n. 警察policeman/policewoman n.男警察/女警察pollution n.污染；污染物population n. 人口，人口数量postcard n. 明信片pound n. 磅；英镑praise v. 表扬；赞扬n. 表扬，赞美prepare v. 是做好准备；把...准备好present n.现在；礼物adj.现在的president n.负责人;主席总统pretty adj. 漂亮的adv. 相当;十分;很pride n. 自愿；骄傲primary adj.最早的;最初的print v.打印；印刷private adj.私人的;私密的probably adv.很可能;大概produce v.生产;制造;出产product n. 产品；制品progress n.&v.进步；进展project n.项目；工程promise n.承诺；诺言v.许诺；承诺pronounce v. 发音pronunciation n.发音proper adj.恰当的;正确的protect v.保护；防护proud adj.自豪的；骄傲的provide v.提供；供应public n.民众adj.公开的；公众的pull v. 拖；拉punish v.处罚；惩罚pupil n.学生purple adj./n.紫色(的) purpose n.目的；目标push v. 鞭策；督促；推动quarter n.一刻钟;四分之一railway n.铁路；铁道raise v.募集；征集rapid adj.迅速的；快速的realise/realize v.理解;领会；意识到receive v.接受;接待;收到recently adv.不久前;最近record n./v.记录refuse v.拒绝regret(v.)懊悔relationship(n.)关系;联系relative亲属;亲戚remain保持不变;剩余repair v. 修理;修补repeat v.重复；重做reply v.回复；回答report v./n. 报道;公布require v.需要;要求research n./v.研究;调查restaurant n. 餐馆return v.归还;回来;返回review v./n.回顾;复习risk n./v.危险;风险;冒险rope n.绳索;绳子v.捆；绑round adj.圆形的rubbish n.垃圾；废弃物Russia n.俄罗斯Russian adj. 俄罗斯的n.俄罗斯人；俄语southern adj.南方的special adj.特别的;特殊的n.特色菜;特价品spirit n.勇气;意志spread v.传播;展开n.蔓延;传播square n.平方;正方形;广场standard n.标准;水平state n.州stick n.棍;条v.粘贴;将…刺入stomach n.胃;腹部stomachache n.胃痛;腹痛straight adj.直的strawberry n.草莓strict adj.严格的;严厉的sugar n.食糖suggest v.建议;提议suggestion n.建议support v./n.支持survey n.调查task n.任务工作technology n.科技;工艺television n.电视机;电视节目temperature n.温度;气温;体温theatre n.戏院;剧场thirsty adj.渴望的;口渴的Thursday n.星期四ticket n.票;入场券tiny adj.极小的;微小的toothache n.牙痛total n.总数;合计adj.总的;全体的toward(s) prep.朝;向;对着translate v.翻译treasure n.珠宝;财富trousers n.裤子truth n.实情,事实Tuesday n.星期二twelfth num.第十二twelve num.十二twentieth num.第二十umbrella n.雨伞university n.大学;高等学府vacation n.假期valuable adj.贵重的;很有用的;宝贵的victory n.胜利;成功video n.录像带;录像village n.村庄;村镇violin n.小提琴violinist n.小提琴手wallet n.钱包warn v.警告;告诫watermelon n.西瓜weather n.天气website n.网站Wednesday n.星期三weight n.重量;份量western adj.西方国的;(尤指)欧美的;西方的wheel n.车轮;轮子whether conj.不管…(还是);或者…(或者)whole adj.䨋全部的整体的winner n.获胜者;优胜者world n.世界worse adj.最差(的);最坏(的);最糟(的)worth adj.值得;有...价值(的)yard n.院子yellow n黄色/adj黄色的zero num.零。
- 1、下载文档前请自行甄别文档内容的完整性，平台不提供额外的编辑、内容补充、找答案等附加服务。
- 2、"仅部分预览"的文档,不可在线预览部分如存在完整性等问题,可反馈申请退款(可完整预览的文档不适用该条件!)。
- 3、如文档侵犯您的权益，请联系客服反馈,我们会尽快为您处理(人工客服工作时间：9:00-18:30)。
1]]11]]]).
%% W a t a s h i h a k o k u s a i s e i j i nitsui.te %% k a k a r e t a hon wo yomu. clinks([ell,]ll],[e12,j13],[e13,j15], [e16,j20]]).
i
If we know the following translation examt)le (2) and (3), we can translate sentence (1) into sentence (4) by imitating examples and colnbining fragments of them. (2) (3)
2
Need to m e nt s
Combine
Frag-
1
Introduction
The basic idea of example-based translation is very simple: translate a source sentence by imitating the translation example of similar sent e n c e i n the database. But in many cases, it is necessary to imitate more than one translation example and combine some fragments of them. Let's consider the translation of the following sentence. (1) He buys a book on international politiasy for a human to do this, but not so for a machine. The ability to combine some fragments of translation examples is essential to example-based translation. A lack of this ability restricts the power of example-based translation. In this paper, we concentrate on the implementation of this ability on machine.
Every ID in an ME should be translatable. We assume the example in Section 3.1 and the following example.
ewd_e( fell, freud,v] , [el2, ['I ),prOn]] , [el3, [book,n] , [el4, [a,det] ] , [elb, Ion,p] , [el6, [politics,n] , felT, [international, adj]
Abstract
An essential problem of example-based translation is how to utilize more than one translation example for translating one source sentence. This 1)aper proposes a method to solve this problem. We introduce tile representation, called .matching e,,z:pressio~z, which tel)resents the combination of fragments of translation examples. The translation process consists of three steps: (.1) Make the source matching expression from lhe source sentence. (2) TransDr the source matching expression into the target matching expression. (3) Construct the target sentence from the target matching expression. This mechanism generates some candidates of translation. To select, the best translation out of them, we define the score of a translation.
ewd e ( [ e l , [ b u y , v ] , [e2,[he,pron]], [e3, [notebook,n], [e4, [a,det]]]]) . %% He buys a notebook. j w d _ e ( [ j I , [kau,v] , [ j 2 , [ha,p] , [j3,[kare,pron]]], [ j 4 , [wo,p] , [ j 5 , [nouto,n]]]]). %% Kare ha nouto wo kau. clinks([[el,jl],[e2,j3],[e3,j5]]). %% el <-> jl, e2 <-> j3, e3 <-> j5
[]
or [<ME-Command> I<ME-Commands>] <ME-Command> : := [d, < ID>] %% delete <ID> or [r,<ID>, <ME>] %% r e p l a c e <ID> %% with <ME> or [a,<ID>,<ME>] %% add <ME> a s a %% c h i l d o f r o o t %% n o d e o f <ID>
Next we will introduce the concept 'matching expression.' Matching expression(ME) is defined as the following:
<HE> : : = [<ID>I<ME-Commands>] <ME-Commands> : :=
inany candidates? We show partial solutions for them in MBT2. MBT2 is the second prototype system in our Memory-based Translation Project.. MBT2 ca.n do bi-directional m~nslation between an English word-dependency tree and a Japanese worddependency tree. It is implemented in Sicstus Prolog.
He buys a notebook. Kate ha nouto wo ka~. I read a boo]~ on international polilics. Watt, hi ha kokusaiseiji nit,suite l:akareta hon wo yomu. Kate ha kokusMseiji nitsuite kM~reta hon
3.1
Translation
Database
The translation database is the collection of translation examples. A t~anslation example consists of three parts: • an English word-dependency tree (EWD) • a Japanese word-dependency tree (JWD) • correspondence links For example, in Prolog,
Toward Memory--based Translation
Satoshi S A T O and Ma.koto N A G A O Dept. of Electrical Engineering, K y o t o University Y o s h i d a - h o n m a c h i , Sa.kyo, K.yoto, 606, Ja.pan sa.to@kuee.kyoto-u.ac.jp
Use of extracted information fiom examples or example-based translation is becoming the new wave of machine translation. The ba.sic idea. of example~based translation is very simple: translate a source sentence by imitating the translation example of a similar sentence in the database. The idea first appeared in [Nagao 84], and some research has followed it [Sumita 88][Sato 89][Sadler 89a.][Sadler 89b]. But a great deal of effort is still needed to implemenl the idea. In our previous work, we show how to select. the best target word in case-frame translation based on examples[Sato 89]. In this paper, we concentrate on two problems: 1. ltow to combine some fragments of translation examph~s in order to translate one sentence? 2. tlow to select tile best tra.nslation out of