外文翻譯---人工智能技術(shù)及應用

1、<p><b>　　中文2150字</b></p><p>　　Technology and Application of Artificial Intelligence</p><p>　　Yao Xifang</p><p>　　AI (Artificial Intelligence), the English abbreviat

2、ion for the AI, is a computer-science, control theory, information theory, linguistics, neurophysiology, psychology, mathematics, philosophy and other subjects of mutual penetration and developed a comprehensive new subj

3、ect . Since its inception through the wave Folding AI, but finally, as an edge of the new discipline recognized by the world and has attracted much interest and concern. Many other disciplines not only introduced or borr

4、owed AI t</p><p>　　Artificial Intelligence of the mankind can be traced back to the seventeenth century Pascal and Leibniz, who had earlier initiation of the idea of ??intelligent machines. Nineteenth centur

5、y, British mathematician and German o Boolean Morgan proposed the "law of thought", which can be described as the beginning of artificial intelligence. Twenties of the nineteenth century, British scientists Bab

6、bage designed the first "computer device", which is considered to be computer hardware, is the predece</p><p>　　As a discipline, artificial intelligence came in 1956, is by the "father of arti

7、ficial intelligence," McCarthy and a group of mathematicians, information scientists, psychologists, neurophysiologists, computer scientists at a meeting of Dartmouth University, the first times proposed. Of artific

8、ial intelligence research, because of the different point of view, the formation of different schools. This is the: symbol school of thought, connecting school of thought school of thought and behavior. </p><p

9、>　　AI is a symbol of traditional doctrine, it Newell and Simon's physical symbol system assumptions. Physical symbol system is an entity composed of a set of symbols, which are the physical model, the entities in

10、the symbolic structure appears as a component can be generated through a variety of operations other symbols structure. Physical symbol system hypothesis: physical symbol system is the intelligent behavior of the suffici

11、ent and necessary conditions. Main task is to "general problem solver"</p><p>　　School of thought is the connection from the human brain structure of the starting system to study the non-procedural

12、, and adaptive, the brain's information processing style, the nature and capacity of neurons in a large number of simple information processing capacity of the Group and its dynamic behavior. </p><p>　　I

13、t is also known as neural computation. Research focus is focused on simulation and realization of man's understanding of the process of sensation, perception process, thinking in images, distributed memory and self-l

14、earning, self-organizing process.</p><p>　　Behaviorist school of psychology, starting from the behavior of that intelligence is only the interaction with the environment in the show. </p><p>　　A

15、rtificial intelligence research has gone through several stages: </p><p>　　The first stage: the rise of artificial intelligence 50 years and cold </p><p>　　After the first time the concept of ar

16、tificial intelligence, there have been a number of significant results, such as machine theorem proving, chess program, general problem solving procedure s, LISP list processing language. However, due to the limited dige

17、stion reasoning, and machine translation, etc. The failure of the artificial intelligence into a low ebb. This stage is characterized by: attention to problem-solving methods, ignoring the importance of knowledge. </p

18、><p>　　The second stage: the late 60s to 70 years, expert systems have enabled a new upsurge of artificial intelligence research.</p><p>　　Chemical mass spectrometry system DENDRAL, MYCIN disease d

19、iagnosis and treatment system, PROSPECTIOR prospecting system, Hearsay-II speech understanding systems, expert systems research and development, will lead a practical artificial intelligence. And in 1969 established the

20、International Joint Conference Artificial Intelligence (International Joint Conferences on Artificial Intelligence is IJCAI). </p><p>　　Phase III: 80 years, with the development of fifth generation computers

21、, artificial intelligence has been great development. </p><p>　　Japan, in 1982 the "fifth generation computer development program," or "knowledge information processing computer system KIPS&qu

22、ot;, its purpose is to make logical reasoning to numerical operations so fast. While this plan ultimately failed, but it formed a study conducted artificial boom. </p><p>　　Phase IV: the late 80s, the rapid

23、development of neural networks. </p><p>　　In 1987, the United States convened the first International Conference on Neural Networks, announced the birth of this new discipline. Since then, the country's

24、investment in the neural network gradually increased, neural network has developed rapidly. </p><p>　　Phase V: 90's, a new study of artificial intelligence climax </p><p>　　As the network te

25、chnology, especially the international development of Internet technology, artificial intelligence research began to shift from the single-agent distributed network environment based on artificial intelligence research.

26、Not only based on the same goal of a distributed problem solving, and studied by the agent for multi-objective problem solving, artificial intelligence is more practical oriented. In addition, the Hopfield neural network

27、 model of the proposed multi-layer, so that</p><p>　　IBM, "Deep Blue" computer beat the human world chess champion, the United States has developed a multi-Agent system application to important res

28、earch information superhighway plan, based on Agent technology Softbot (software robots) in software and network search engines has been fully applied at the same time, the U.S. Sandia Laboratories has established the In

29、ternational's largest "virtual reality" laboratory, to be adopted by the helmet and data glove data to achieve a more friendly human-com</p><p>　　The main content of current artificial intellig

30、ence: distributed artificial intelligence and multi-agent system, artificial model of thinking, knowledge systems (including expert systems, knowledge-based systems and intelligent decision system), Knowledge Discovery a

31、nd Data Mining (from a large number of incomplete , fuzzy, noisy data mining useful knowledge out of us), Genetic and Evolutionary Computation (Genetic and Evolutionary Biology by theoretical simulations to reveal the ev

32、olutionary la</p><p>　　Although research and application of artificial intelligence made a lot of results, but there is a great application from the full distance, there are many issues to be resolved, and t

33、he need for a multidisciplinary research experts work together. Future research directions in artificial intelligence are: artificial intelligence, machine learning models and theories, imprecise knowledge representation

34、 and reasoning, common sense knowledge and reasoning, artificial model of thinking, intelligent </p><p><b>　　人工智能技術(shù)及應用</b></p><p><b>　　姚錫房</b></p><p>　　人工智能（

35、Artificial Intelligence）, 英文縮寫為AI,是一門由計算機科學、控制論、信息論、語言學、神經(jīng)生理學、心理學、數(shù)學、哲學等多種學科相互滲透而發(fā)展起來的綜合性新學科。自問世以來AI經(jīng)過波波折折，但終于作為一門邊緣新學科得到世界的承認并且日益引起人們的興趣和關(guān)注。不僅許多其他學科開始引入或借用AI技術(shù)，而且AI中的專家系統(tǒng)、自然語言處理和圖象識別已成為新興的知識產(chǎn)業(yè)的三大突破口。 <

36、/p><p>　　人工智能的思想萌芽可以追溯到十七世紀的巴斯卡和萊布尼茨，他們較早萌生了有智能的機器的想法。十九世紀，英國數(shù)學家布爾和德o摩爾根提出了“思維定律“，這些可謂是人工智能的開端。十九世紀二十年代，英國科學家巴貝奇設(shè)計了第一架“計算機器“，它被認為是計算機硬件，也是人工智能硬件的前身。電子計算機的問世，使人工智能的研究真正成為可能。 </p><p>　　作為一門學科，人工

37、智能于1956年問世，是由“人工智能之父“McCarthy及一批數(shù)學家、信息學家、心理學家、神經(jīng)生理學家、計算機科學家在Dartmouth大學召開的會議上，首次提出。對人工智能的研究，由于研究角度的不同，形成了不同的研究學派。這就是：符號主義學派、連接主義學派和行為主義學派。 </p><p>　　傳統(tǒng)人工智能是符號主義，它以Newell和Simon提出的物理符號系統(tǒng)假設(shè)為基礎(chǔ)。物理符號系統(tǒng)是由一組符號

38、實體組成，它們都是物理模式，可在符號結(jié)構(gòu)的實體中作為組成成分出現(xiàn)，可通過各種操作生成其它符號結(jié)構(gòu)。物理符號系統(tǒng)假設(shè)認為：物理符號系統(tǒng)是智能行為的充分和必要條件。主要工作是“通用問題求解程序”：通過抽象，將一個現(xiàn)實系統(tǒng)變成一個符號系統(tǒng)，基于此符號系統(tǒng)，使用動態(tài)搜索方法求解問題。</p><p>　　連接主義學派是從人的大腦神經(jīng)系統(tǒng)結(jié)構(gòu)出發(fā)，研究非程序的、適應性的、大腦風格的信息處理的本質(zhì)和能力，研究大量簡單的神經(jīng)元

39、的集團信息處理能力及其動態(tài)行為。</p><p>　　人們也稱之為神經(jīng)計算。研究重點是側(cè)重于模擬和實現(xiàn)人的認識過程中的感覺、知覺過程、形象思維、分布式記憶和自學習、自組織過程。 </p><p>　　行為主義學派是從行為心理學出發(fā)，認為智能只是在與環(huán)境的交互作用中表現(xiàn)出來。</p><p>　　人工智能的研究經(jīng)歷了以下幾個階段： </p&g

40、t;<p>　　第一階段：50年代人工智能的興起和冷落 </p><p>　　人工智能概念首次提出后，相繼出現(xiàn)了一批顯著的成果，如機器定理證明、跳棋程序、通用問題s求解程序、LISP表處理語言等。但由于消解法推理能力的有限，以及機器翻譯等的失敗，使人工智能走入了低谷。這一階段的特點是：重視問題求解的方法，忽視知識重要性。 </p><p>　　第二階段：6

41、0年代末到70年代，專家系統(tǒng)出現(xiàn)，使人工智能研究出現(xiàn)新高潮 </p><p>　　DENDRAL化學質(zhì)譜分析系統(tǒng)、MYCIN疾病診斷和治療系統(tǒng)、PROSPECTIOR探礦系統(tǒng)、Hearsay-II語音理解系統(tǒng)等專家系統(tǒng)的研究和開發(fā)，將人工智能引向了實用化。并且，1969年成立了國際人工智能聯(lián)合會議。</p><p>　　第三階段：80年代，隨著第五代計算機的研制，人工智能得到了很

42、大發(fā)展 </p><p>　　日本1982年開始了“第五代計算機研制計劃“，即“知識信息處理計算機系統(tǒng)KIPS“，其目的是使邏輯推理達到數(shù)值運算那么快。雖然此計劃最終失敗，但它的開展形成了一股研究人工智能的熱潮。 </p><p>　　第四階段：80年代末，神經(jīng)網(wǎng)絡(luò)飛速發(fā)展 </p><p>　　1987年，美國召開第一次神經(jīng)網(wǎng)絡(luò)國際會

43、議，宣告了這一新學科的誕生。此后，各國在神經(jīng)網(wǎng)絡(luò)方面的投資逐漸增加，神經(jīng)網(wǎng)絡(luò)迅速發(fā)展起來。 </p><p>　　第五階段：90年代，人工智能出現(xiàn)新的研究高潮 </p><p>　　由于網(wǎng)絡(luò)技術(shù)特別是國際互連網(wǎng)的技術(shù)發(fā)展，人工智能開始由單個智能主體研究轉(zhuǎn)向基于網(wǎng)絡(luò)環(huán)境下的分布式人工智能研究。不僅研究基于同一目標的分布式問題求解，而且研究多個智能主體的多目標問題求解，將人

44、工智能更面向?qū)嵱?。另外，由于Hopfield多層神經(jīng)網(wǎng)絡(luò)模型的提出，使人工神經(jīng)網(wǎng)絡(luò)研究與應用出現(xiàn)了欣欣向榮的景象。人工智能已深入到社會生活的各個領(lǐng)域。</p><p>　?。桑拢凸尽吧钏{“電腦擊敗了人類的世界國際象棋冠軍，美國制定了以多Agent系統(tǒng)應用為重要研究內(nèi)容的信息高速公路計劃，基于Agent技術(shù)的Softbot(軟機器人)在軟件領(lǐng)域和網(wǎng)絡(luò)搜索引擎中得到了充分應用，同時，美國Sandia實驗室建立了國際

45、上最龐大的“虛擬現(xiàn)實“實驗室，擬通過數(shù)據(jù)頭盔和數(shù)據(jù)手套實現(xiàn)更友好的人機交互，建立更好的智能用戶接口。圖像處理和圖像識別，聲音處理和聲音識別取得了較好的發(fā)展，IBM公司推出了ViaVoice聲音識別軟件，以使聲音作為重要的信息輸入媒體。國際各大計算機公司又開始將“人工智能“作為其研究內(nèi)容。人們普遍認為，計算機將會向網(wǎng)絡(luò)化、智能化、并行化方向發(fā)展。二十一世紀的信息技術(shù)領(lǐng)域?qū)灾悄苄畔⑻幚頌橹行摹?lt;/p><p>　

46、　目前人工智能主要研究內(nèi)容是：分布式人工智能與多智能主體系統(tǒng)、人工思維模型、知識系統(tǒng)（包括專家系統(tǒng)、知識庫系統(tǒng)和智能決策系統(tǒng)）、知識發(fā)現(xiàn)與數(shù)據(jù)挖掘（從大量的、不完全的、模糊的、有噪聲的數(shù)據(jù)中挖掘出對我們有用的知識）、遺傳與演化計算（通過對生物遺傳與進化理論的模擬，揭示出人的智能進化規(guī)律）、人工生命（通過構(gòu)造簡單的人工生命系統(tǒng)（如：機器蟲）并觀察其行為，探討初級智能的奧秘）、人工智能應用（如：模糊控制、智能大廈、智能人機接口、智能機器人等