外文翻譯-----數(shù)據(jù)挖掘什么是數(shù)據(jù)挖掘？

1、<p>　　畢業(yè)設(shè)計(jì)（論文）外文翻譯</p><p>　　——基于數(shù)據(jù)挖掘技術(shù)的WWW推薦系統(tǒng)設(shè)計(jì)</p><p><b>　　英文原文</b></p><p>　　Data Mining: What is Data Mining?</p><p><b>　　Overview </b>&

2、lt;/p><p>　　Generally, data mining (sometimes called data or knowledge discovery) is the process of analyzing data from different perspectives and summarizing it into useful information - information that can b

3、e used to increase revenue, cuts costs, or both. Data mining software is one of a number of analytical tools for analyzing data. It allows users to analyze data from many different dimensions or angles, categorize it, an

4、d summarize the relationships identified. Technically, data mining is the process </p><p>　　Continuous Innovation </p><p>　　Although data mining is a relatively new term, the technology is not.

5、Companies have used powerful computers to sift through volumes of supermarket scanner data and analyze market research reports for years. However, continuous innovations in computer processing power, disk storage, and st

6、atistical software are dramatically increasing the accuracy of analysis while driving down the cost. </p><p><b>　　Example </b></p><p>　　For example, one Midwest grocery chain used th

7、e data mining capacity of Oracle software to analyze local buying patterns. They discovered that when men bought diapers on Thursdays and Saturdays, they also tended to buy beer. Further analysis showed that these shoppe

8、rs typically did their weekly grocery shopping on Saturdays. On Thursdays, however, they only bought a few items. The retailer concluded that they purchased the beer to have it available for the upcoming weekend. The gro

9、cery chain cou</p><p>　　Data, Information, and Knowledge </p><p><b>　　Data</b></p><p>　　Data are any facts, numbers, or text that can be processed by a computer. Today,

10、organizations are accumulating vast and growing amounts of data in different formats and different databases. This includes: </p><p>　　?operational or transactional data such as, sales, cost, inventory, pay

11、roll, and accounting</p><p>　　?nonoperational data, such as industry sales, forecast data, and macro economic data </p><p>　　?meta data - data about the data itself, such as logical database d

12、esign or data dictionary definitions </p><p>　　Information</p><p>　　The patterns, associations, or relationships among all this data can provide information. For example, analysis of retail poin

13、t of sale transaction data can yield information on which products are selling and when. </p><p><b>　　Knowledge</b></p><p>　　Information can be converted into knowledge about histori

14、cal patterns and future trends. For example, summary information on retail supermarket sales can be analyzed in light of promotional efforts to provide knowledge of consumer buying behavior. Thus, a manufacturer or retai

15、ler could determine which items are most susceptible to promotional efforts. </p><p>　　Data Warehouses </p><p>　　Dramatic advances in data capture, processing power, data transmission, and stora

16、ge capabilities are enabling organizations to integrate their various databases into data warehouses. Data warehousing is defined as a process of centralized data management and retrieval. Data warehousing, like data min

17、ing, is a relatively new term although the concept itself has been around for years. Data warehousing represents an ideal vision of maintaining a central repository of all organizational data. Centra</p><p>

18、　　What can data mining do? </p><p>　　Data mining is primarily used today by companies with a strong consumer focus - retail, financial, communication, and marketing organizations. It enables these companies

19、to determine relationships among "internal" factors such as price, product positioning, or staff skills, and "external" factors such as economic indicators, competition, and customer demographics. And

20、, it enables them to determine the impact on sales, customer satisfaction, and corporate profits. Finally, it enables them to "dri</p><p>　　With data mining, a retailer could use point-of-sale records o

21、f customer purchases to send targeted promotions based on an individual's purchase history. By mining demographic data from comment or warranty cards, the retailer could develop products and promotions to appeal to s

22、pecific customer segments. </p><p>　　For example, Blockbuster Entertainment mines its video rental history database to recommend rentals to individual customers. American Express can suggest products to its

23、cardholders based on analysis of their monthly expenditures. </p><p>　　WalMart is pioneering massive data mining to transform its supplier relationships. WalMart captures point-of-sale transactions from over

24、 2,900 stores in 6 countries and continuously transmits this data to its massive 7.5 terabyte Teradata data warehouse. WalMart allows more than 3,500 suppliers, to access data on their products and perform data analyses.

25、 These suppliers use this data to identify customer buying patterns at the store display level. They use this information to manage local store </p><p>　　The National Basketball Association (NBA) is explorin

26、g a data mining application that can be used in conjunction with image recordings of basketball games. The Advanced Scout software analyzes the movements of players to help coaches orchestrate plays and strategies. For e

27、xample, an analysis of the play-by-play sheet of the game played between the New York Knicks and the Cleveland Cavaliers on January 6, 1995 reveals that when Mark Price played the Guard position, John Williams attempted

28、four ju</p><p>　　By using the NBA universal clock, a coach can automatically bring up the video clips showing each of the jump shots attempted by Williams with Price on the floor, without needing to comb thr

29、ough hours of video footage. Those clips show a very successful pick-and-roll play in which Price draws the Knick's defense and then finds Williams for an open jump shot. </p><p>　　How does data mining w

30、ork? </p><p>　　While large-scale information technology has been evolving separate transaction and analytical systems, data mining provides the link between the two. Data mining software analyzes relationshi

31、ps and patterns in stored transaction data based on open-ended user queries. Several types of analytical software are available: statistical, machine learning, and neural networks. Generally, any of four types of relatio

32、nships are sought: </p><p>　　?Classes: Stored data is used to locate data in predetermined groups. For example, a restaurant chain could mine customer purchase data to determine when customers visit and wha

33、t they typically order. This information could be used to increase traffic by having daily specials.</p><p>　　?Clusters: Data items are grouped according to logical relationships or consumer preferences. Fo

34、r example, data can be mined to identify market segments or consumer affinities. </p><p>　　?Associations: Data can be mined to identify associations. The beer-diaper example is an example of associative min

35、ing. </p><p>　　?Sequential patterns: Data is mined to anticipate behavior patterns and trends. For example, an outdoor equipment retailer could predict the likelihood of a backpack being purchased based on

36、a consumer's purchase of sleeping bags and hiking shoes. </p><p>　　Data mining consists of five major elements: </p><p>　　?Extract, transform, and load transaction data onto the data wareho

37、use system. </p><p>　　?Store and manage the data in a multidimensional database system. </p><p>　　?Provide data access to business analysts and information technology professionals. </p>

38、<p>　　?Analyze the data by application software. </p><p>　　?Present the data in a useful format, such as a graph or table. </p><p>　　Different levels of analysis are available: </p&g

39、t;<p>　　?Artificial neural networks: Non-linear predictive models that learn through training and resemble biological neural networks in structure.</p><p>　　?Genetic algorithms: Optimization techniq

40、ues that use processes such as genetic combination, mutation, and natural selection in a design based on the concepts of natural evolution. </p><p>　　?Decision trees: Tree-shaped structures that represent s

41、ets of decisions. These decisions generate rules for the classification of a dataset. Specific decision tree methods include Classification and Regression Trees (CART) and Chi Square Automatic Interaction Detection (CHAI

42、D) . CART and CHAID are decision tree techniques used for classification of a dataset. They provide a set of rules that you can apply to a new (unclassified) dataset to predict which records will have a given outcome. CA

43、RT </p><p>　　?Nearest neighbor method: A technique that classifies each record in a dataset based on a combination of the classes of the k record(s) most similar to it in a historical dataset (where k 1). S

44、ometimes called the k-nearest neighbor technique. </p><p>　　?Rule induction: The extraction of useful if-then rules from data based on statistical significance. </p><p>　　?Data visualization:

45、The visual interpretation of complex relationships in multidimensional data. Graphics tools are used to illustrate data relationships. </p><p>　　What technological infrastructure is required?</p><

46、p>　　Today, data mining applications are available on all size systems for mainframe, client/server, and PC platforms. System prices range from several thousand dollars for the smallest applications up to $1 million a

47、terabyte for the largest. Enterprise-wide applications generally range in size from 10 gigabytes to over 11 terabytes. NCR has the capacity to deliver applications exceeding 100 terabytes. There are two critical technolo

48、gical drivers: </p><p>　　?Size of the database: the more data being processed and maintained, the more powerful the system required. </p><p>　　?Query complexity: the more complex the queries a

49、nd the greater the number of queries being processed, the more powerful the system required. </p><p>　　Relational database storage and management technology is adequate for many data mining applications less

50、 than 50 gigabytes. However, this infrastructure needs to be significantly enhanced to support larger applications. Some vendors have added extensive indexing capabilities to improve query performance. Others use new har

51、dware architectures such as Massively Parallel Processors (MPP) to achieve order-of-magnitude improvements in query time. For example, MPP systems from NCR link hundreds of hig</p><p><b>　　中文部分</b&g

52、t;</p><p>　　數(shù)據(jù)挖掘：什么是數(shù)據(jù)挖掘？</p><p><b>　　概述</b></p><p>　　一般來(lái)說(shuō)，數(shù)據(jù)挖掘（有時(shí)也被稱為數(shù)據(jù)或知識(shí)發(fā)現(xiàn)）是從不同的角度進(jìn)行分析和總結(jié)數(shù)據(jù)，進(jìn)而轉(zhuǎn)化為有用信息的IT流程，這些信息可以用于增加收入，降低成本，或兩者兼而有之。數(shù)據(jù)挖掘軟件是用來(lái)分析數(shù)據(jù)的工具之一。它允許用戶分析來(lái)自許多不同的

53、層面或角度的數(shù)據(jù)，歸類、發(fā)現(xiàn)和總結(jié)其中的關(guān)系。從技術(shù)上講，數(shù)據(jù)挖掘是發(fā)現(xiàn)在大型關(guān)系數(shù)據(jù)庫(kù)中的數(shù)十個(gè)相關(guān)領(lǐng)域或模式的過(guò)程。</p><p><b>　　不斷創(chuàng)新</b></p><p>　　雖然數(shù)據(jù)挖掘是一個(gè)相對(duì)較新的術(shù)語(yǔ)，但從技術(shù)上來(lái)說(shuō)并非如此。在以前，公司主要通過(guò)強(qiáng)大的計(jì)算機(jī)來(lái)篩選超市掃描數(shù)據(jù)量和多年的市場(chǎng)分析研究報(bào)告來(lái)進(jìn)行數(shù)據(jù)分析?，F(xiàn)在，計(jì)算機(jī)的處理能力，磁盤存儲(chǔ)

54、和統(tǒng)計(jì)軟件的不斷創(chuàng)新，都正在顯著的提高分析的準(zhǔn)確性，同時(shí)降低成本。</p><p><b>　　范例</b></p><p>　　例如，一個(gè)中西部雜貨連鎖店使用Oracle軟件的數(shù)據(jù)挖掘能力，以分析當(dāng)?shù)氐馁?gòu)買模式。他們發(fā)現(xiàn)，男子在周四和周六買尿布的時(shí)候，他們也傾向于購(gòu)買啤酒。進(jìn)一步分析表明，這些購(gòu)物者通常因?yàn)橐习喽粫?huì)在星期六買菜。但是，星期四他們也只買了幾種商品。

55、這家零售商認(rèn)為他們購(gòu)買的啤酒主要用于即將到來(lái)的周末。雜貨連鎖店可以利用這個(gè)新發(fā)現(xiàn)的購(gòu)物方式信息來(lái)增加收入。例如，他們可以將啤酒和尿布擺在一起。而且，他們可以確保啤酒和尿布是在上周四全價(jià)出售。</p><p><b>　　數(shù)據(jù)，信息和知識(shí)</b></p><p><b>　　數(shù)據(jù)</b></p><p>　　數(shù)據(jù)是任何事實(shí)，

56、數(shù)字或文字，可以由計(jì)算機(jī)處理。今天，企業(yè)也積累了在不同數(shù)據(jù)庫(kù)中的數(shù)據(jù)格式和不同的廣闊和不斷增長(zhǎng)的數(shù)額。這包括：</p><p>　　?業(yè)務(wù)或交易數(shù)據(jù)，例如，銷售，成本，庫(kù)存，工資和會(huì)計(jì)</p><p>　　?nonoperational數(shù)據(jù)，如行業(yè)銷售，預(yù)測(cè)數(shù)據(jù)和宏觀經(jīng)濟(jì)數(shù)據(jù)</p><p>　　?元數(shù)據(jù) - 關(guān)于數(shù)據(jù)本身的數(shù)據(jù)，如數(shù)據(jù)庫(kù)的邏輯設(shè)計(jì)或數(shù)據(jù)字典的

57、定義</p><p><b>　　信息</b></p><p>　　模式，關(guān)聯(lián)，或在所有這些數(shù)據(jù)之間的關(guān)系都可以提供信息。例如，銷售交易數(shù)據(jù)分析，零售點(diǎn)的信息可以產(chǎn)生哪些產(chǎn)品銷售和時(shí)間。</p><p><b>　　知識(shí)</b></p><p>　　信息可以被轉(zhuǎn)化為對(duì)歷史規(guī)律和未來(lái)趨勢(shì)的了解。例如，

58、總結(jié)零售超市銷售信息可以分析推廣工作，提供消費(fèi)者購(gòu)買行為的光。因此，制造商或零售商可以決定哪些東西最容易推廣。</p><p><b>　　數(shù)據(jù)倉(cāng)庫(kù)</b></p><p>　　迅猛發(fā)展的數(shù)據(jù)采集，處理能力，數(shù)據(jù)傳輸和存儲(chǔ)能力使企業(yè)能夠整合數(shù)據(jù)倉(cāng)庫(kù)的各種數(shù)據(jù)庫(kù)。數(shù)據(jù)倉(cāng)庫(kù)是指一個(gè)集中的數(shù)據(jù)管理和檢索的過(guò)程。數(shù)據(jù)倉(cāng)庫(kù)，和數(shù)據(jù)挖掘一樣，雖然本身是一個(gè)已經(jīng)存在多年的概念相對(duì)較

59、新的任期。數(shù)據(jù)倉(cāng)庫(kù)代表了維護(hù)所有組織數(shù)據(jù)的中央儲(chǔ)存庫(kù)的理想目標(biāo)。數(shù)據(jù)集中是需要最大化的用戶訪問(wèn)和分析。戲劇性的技術(shù)進(jìn)步使這一設(shè)想成為許多企業(yè)的現(xiàn)實(shí)。而且，在數(shù)據(jù)分析軟件同樣巨大的進(jìn)步使用戶能夠自由地訪問(wèn)該數(shù)據(jù)。數(shù)據(jù)分析軟件是支持?jǐn)?shù)據(jù)挖掘。</p><p>　　數(shù)據(jù)挖掘可以做什么？</p><p>　　主要用于數(shù)據(jù)挖掘的公司今天在密切關(guān)注消費(fèi)者 - 零售，金融，通信和營(yíng)銷組織。它使這些公司來(lái)

60、決定在“內(nèi)部”，如價(jià)格，產(chǎn)品定位，或工作人員的技能因素的關(guān)系，和“外部”，如經(jīng)濟(jì)指標(biāo)，競(jìng)爭(zhēng)和客戶的人口統(tǒng)計(jì)因素。而且，它使他們能夠確定在銷售，客戶滿意度和企業(yè)利潤(rùn)的影響。最后，它使他們能夠“深入”到摘要信息，查看詳細(xì)的交易數(shù)據(jù)。</p><p>　　通過(guò)數(shù)據(jù)挖掘，零售商可以使用的客戶購(gòu)買點(diǎn)的銷售記錄，發(fā)送個(gè)人的購(gòu)買記錄為基礎(chǔ)針對(duì)性的促銷。通過(guò)挖掘意見或保修卡從人口統(tǒng)計(jì)數(shù)據(jù)，零售商可以開發(fā)產(chǎn)品和促銷活動(dòng)，吸引特定的

61、客戶群。</p><p>　　例如，百視達(dá)娛樂(lè)地雷的錄影帶出租租金歷史數(shù)據(jù)庫(kù)中，建議對(duì)個(gè)人客戶。美國(guó)運(yùn)通持卡人可以建議其產(chǎn)品的基礎(chǔ)上，他們每月支出的分析。</p><p>　　沃爾瑪是開拓龐大的數(shù)據(jù)挖掘改變其供應(yīng)商關(guān)系。沃爾瑪捕捉來(lái)自6個(gè)國(guó)家的2,900點(diǎn)店的銷售交易，并不斷傳遞到其龐大的7.5 TB的Teradata數(shù)據(jù)倉(cāng)庫(kù)的數(shù)據(jù)。沃爾瑪允許超過(guò)3500個(gè)供應(yīng)商，其產(chǎn)品上訪問(wèn)數(shù)據(jù)和執(zhí)行數(shù)

62、據(jù)分析。這些供應(yīng)商使用這些數(shù)據(jù)來(lái)確定在店內(nèi)展示級(jí)客戶購(gòu)買模式。他們利用這些信息來(lái)管理本地商店庫(kù)存，并確定新的銷售機(jī)會(huì)。 1995年，沃爾瑪計(jì)算機(jī)處理超過(guò)100萬(wàn)的查詢復(fù)雜數(shù)據(jù)。</p><p>　　美國(guó)國(guó)家籃球協(xié)會(huì)（NBA）是數(shù)據(jù)挖掘中的應(yīng)用探索，可配合使用的籃球比賽的影像記錄。高級(jí)軟件分析球員的動(dòng)作來(lái)幫助教練編排戰(zhàn)術(shù)和策略。例如，分析的播放按播放之間的紐約尼克斯隊(duì)和克利夫蘭騎士隊(duì)1月6日起在游戲片，1995年發(fā)

63、現(xiàn)時(shí)，馬克后衛(wèi)的位置上發(fā)揮的價(jià)值，約翰威廉姆斯企圖四跳投，并提出各一！先進(jìn)的偵察兵，不僅認(rèn)為這種模式，但解釋說(shuō)這是相當(dāng)有趣，因?yàn)樗煌谝话闩臄z從49.30％的比例在騎士隊(duì)那場(chǎng)比賽。</p><p>　　利用NBA的通用時(shí)鐘，一個(gè)教練可以自動(dòng)彈出，而無(wú)需通過(guò)梳理小時(shí)的錄像短片顯示威廉姆斯試圖用價(jià)格在地板上的每一個(gè)跳投。這些剪輯顯示一個(gè)非常成功的挑選和角色扮演，其中價(jià)格提請(qǐng)尼克斯隊(duì)的防守，然后找到一個(gè)開放的跳投威廉

64、姆斯。</p><p>　　數(shù)據(jù)挖掘是如何工作的？</p><p>　　盡管大型信息技術(shù)已經(jīng)發(fā)展獨(dú)立的交易和分析系統(tǒng)，數(shù)據(jù)挖掘提供了兩者之間的聯(lián)系。數(shù)據(jù)挖掘軟件分析在存儲(chǔ)交易開放式的用戶查詢的數(shù)據(jù)關(guān)系和模式。分析軟件提供了幾種類型：統(tǒng)計(jì)，機(jī)器學(xué)習(xí)，神經(jīng)網(wǎng)絡(luò)。一般來(lái)說(shuō)，四種類型的關(guān)系的任何要求：</p><p>　　類：存儲(chǔ)的數(shù)據(jù)是用來(lái)定位在預(yù)定的組的數(shù)據(jù)。例如，我的

65、餐飲連鎖企業(yè)客戶可以購(gòu)買數(shù)據(jù)，以確定哪些客戶光臨時(shí)，他們通常的順序。此信息可能被用來(lái)增加每天有特價(jià)流量。</p><p>　　集群：數(shù)據(jù)項(xiàng)進(jìn)行分組根據(jù)邏輯關(guān)系或消費(fèi)者的喜好。例如，可以將數(shù)據(jù)挖掘，以確定細(xì)分市場(chǎng)或消費(fèi)者的親和力。</p><p>　　社團(tuán)：數(shù)據(jù)可以開采，以確定關(guān)聯(lián)。啤酒，尿布的例子是一個(gè)關(guān)聯(lián)挖掘的例子。</p><p>　　序列模式：數(shù)據(jù)挖掘，預(yù)測(cè)行

66、為模式和趨勢(shì)。例如，一個(gè)戶外設(shè)備零售商可以預(yù)測(cè)的可能性的背包上購(gòu)買消費(fèi)者的睡袋，登山鞋購(gòu)買基礎(chǔ)。</p><p>　　數(shù)據(jù)挖掘包括五個(gè)主要元素：</p><p>　　提取，轉(zhuǎn)換和交易數(shù)據(jù)加載到數(shù)據(jù)倉(cāng)庫(kù)系統(tǒng)。</p><p>　　存儲(chǔ)和管理多維數(shù)據(jù)庫(kù)系統(tǒng)的數(shù)據(jù)。</p><p>　　提供數(shù)據(jù)訪問(wèn)業(yè)務(wù)分析員和信息技術(shù)專業(yè)人才。</p>

67、<p>　　由應(yīng)用軟件分析數(shù)據(jù)。</p><p>　　目前在一個(gè)有用的格式的數(shù)據(jù)，如圖形或表。</p><p><b>　　不同的分析級(jí)別：</b></p><p>　　人工神經(jīng)網(wǎng)絡(luò)：非線性是通過(guò)培訓(xùn)和學(xué)習(xí)的結(jié)構(gòu)類似生物神經(jīng)網(wǎng)絡(luò)預(yù)測(cè)模型。</p><p>　　遺傳算法：使用優(yōu)化技術(shù)的結(jié)合過(guò)程，如遺傳，變異，

68、并在對(duì)自然進(jìn)化的概念為基礎(chǔ)設(shè)計(jì)的自然選擇。</p><p>　　決策樹：樹狀結(jié)構(gòu)代表的決定套。這些決定產(chǎn)生的數(shù)據(jù)集分類規(guī)則。具體方法包括決策樹分類回歸樹（CART）的和卡方自動(dòng)交互檢測(cè)（CHAID）。 CART和CHAID的決策樹分類數(shù)據(jù)集使用的技術(shù)。它們提供了一種規(guī)則，你可以申請(qǐng)一個(gè)新的（未分類）數(shù)據(jù)集以預(yù)測(cè)哪些記錄將有相應(yīng)的結(jié)果集。通過(guò)創(chuàng)建車段2路數(shù)據(jù)集分割而CHAID以卡方檢定段創(chuàng)建多路分割。車通常需要不到

69、CHAID數(shù)據(jù)準(zhǔn)備。</p><p>　　最近鄰法：一個(gè)技術(shù)，在數(shù)據(jù)集分類的基礎(chǔ)上每個(gè)記錄的記錄的K（s）最相似的歷史數(shù)據(jù)集給它的類的組合（其中k 1）。有時(shí)被稱為的K -最近鄰技術(shù)。</p><p>　　規(guī)則歸納：運(yùn)用有用的if - then規(guī)則從統(tǒng)計(jì)意義的數(shù)據(jù)提取。</p><p>　　數(shù)據(jù)可視化：在多維數(shù)據(jù)的復(fù)雜關(guān)系的可視化解釋。圖形工具是用來(lái)說(shuō)明數(shù)據(jù)關(guān)系。&

70、lt;/p><p>　　科技基礎(chǔ)設(shè)施是需要什么？</p><p>　　如今，數(shù)據(jù)挖掘應(yīng)用，可為主機(jī)，客戶機(jī)/服務(wù)器，PC平臺(tái)的所有大小的系統(tǒng)。系統(tǒng)的價(jià)格范圍從幾千年的最小應(yīng)用的最大上限為100萬(wàn)TB的美元。企業(yè)范圍的應(yīng)用，一般的尺寸范圍從10千兆字節(jié)到超過(guò)11萬(wàn)億字節(jié)。 NCR已交付應(yīng)用的能力超過(guò)100千兆字節(jié)。有兩個(gè)關(guān)鍵的技術(shù)驅(qū)動(dòng)程序：</p><p>　　大小數(shù)據(jù)庫(kù)

71、：更多數(shù)據(jù)正在處理和維護(hù)，更強(qiáng)大的系統(tǒng)需要。</p><p>　　查詢的復(fù)雜性：更復(fù)雜的查詢和查詢的數(shù)量越大，正在處理，更強(qiáng)大的系統(tǒng)需要。</p><p>　　關(guān)系型數(shù)據(jù)庫(kù)存儲(chǔ)和管理技術(shù)是多種數(shù)據(jù)挖掘應(yīng)用超過(guò)50千兆字節(jié)更難滿足。但是，這種基礎(chǔ)設(shè)施的需求將明顯增強(qiáng)，以支持更大的應(yīng)用程序。一些廠商已經(jīng)將廣泛索引功能來(lái)提高查詢性能。其他使用新的硬件，如大量的并行處理器架構(gòu)（MPP）的實(shí)現(xiàn)訂單在