1、专业英语研究生期末考试四川大学研究生考试样题四川大学计算机（软件）学院工程硕士考试试题（20122013学年上学期）课程号：_课序号：_课程名称： 专业英语 任课教师：于中华 适用专业： 软件工程 适用年级：_学生人数：_印题份数：_学号：_姓名：_考 试 须 知四川大学学生参加由学校组织或由学院承办的各级各类考试，必须严格执行四川大学考试工作管理办法和四川大学考场规则。有考试违纪作弊行为的，一律按照四川大学学生考试违纪作弊处罚规定进行处理。四川大学各级各类考试的监考人员，必须严格执行四川大学考试工作管理办法、四川大学考场规则和四川大学监考人员职责。有违反学校有关规定的，严格按照四川大教学事故
2、认定及处理办法进行处理。题号一二三四五六七八九十平时总分得分考试时间 年 月 日阅卷教师签名I. Choose the best answer to each of the following questions. (10%)1._ is used to communicate with another computer over telephone lines A. keyboard B .modem C. printer D. mouse2._is a device that enables the computer to handle sounds.A. network card B.
3、video card C. sound card D. monitor3. _ refers to the parts of the computer that you can see and touch.A. Software B. Hardware C. Hardship D. Instruction4. The display screen is the most common device used to show you what the computer is doing.A. input B. printing C. output D. electronic5. _ are pr
4、ograms that provide access to Web resources A. browsers B. database C. URL D. E-commerce6._ is a word processing tool that helps you to create quality documents.A. Word 2003 B. Outlook 2003 C. Access 2003 D. Excel 20037. _ is a fast and powerful operating system, based on the UNIX O.S.A. Windows 200
5、3 B. Office 2003 C. Windows 2003 D. Linux8. A virus is a _A. program B. computer C. bad man D. beast9. TCP/IP are the two standard protocols for communications on the _. A. web B. computer C. Internet D. network10. B2C involves individuals selling to _. A. client B. consumer C. corporation D. compan
6、yII. Decide whether the following statements are True or False. (10%)1. ( ) The CPU is the processor of a computer. 2. ( ) Windows NT is designed for replacement of Windows.3. ( ) A computer is only connected by cable lines, such as telephone lines.4. ( ) Air can be used to transit data.5. ( ) Searc
7、h engines help you to locate information on the Web6. ( ) B2C involves individuals selling to individuals.7. ( ) Databases are widely used8. ( ) Oracle is the largest database system vendor in the world.9. ( ) Complete access and complete security are good for information security10. ( ) Portable co
8、mputer can fit in a briefcase or even in the palm of your hand.III. Translation: English to Chinese (20%) 1. CPU_ 2. OS _ 3. LAN_ 4. B2B_ 5. CEO_ 6. ATM_ 7. DBMS_ 8. URL_ 9. personal computer _10. storage device _11. information exchanges _12. protocol _13. broadband _14.browser _15. search engine _
9、16. update _17.Web Site _18. human resources _19. environment _20. computer virus _IV. Word-matching (10%)1. LCD a. 主页2. menus b. 黑客3. bus c.超链接4. WAN d. 冲浪 5. home page e液晶显示器6. E-commerce f 菜单7. hacker g 总线8. surfing h 广域网9. delete i电子商务10. hyperlink j 删除VTranslation into Chinese (20%) 1. Computer
10、 hardware has four parts: the central processing unit (CPU) and memory, storage hardware, input hardware, and output hardware. 2. Software refers to computer programs. 3. Computer network is a system connecting two or more computers. 4. Communication is the most popular Internet activity 5. The back
11、bone of this electronic commerce is the Internet. 6. DBMS has the ability to access large amounts of data efficiently. 7. Security is important throughout the information life-cycle 8. If you have access to a personal computer (PC) and can connect to the Internet with a browser, you can do business
12、online. 9. Business receives raw materials as input, add value to them through various processes, and sell the finished product as output to customers. 10. As new viruses are created everyday, upgrade your anti-virus software regularly. VTranslation into English (30%)1. 随着中国电子商务的迅速发展，消费者在购物网站上留下的产品评
13、论越来越多。通过参考这些评论，消费者可以决定是否购买某件产品，生产商也可以对产品进行改进，因此产品评论挖掘的研究具有重要意义。而产品特征抽取是产品评论挖掘的基础，直接影响到产品评论挖掘的结果。本文面向中文电子商务网站的用户评论，针对现有方法移植性差，准确率不高的缺点，提出了一种产品特征的自动抽取方法。该方法首先使用词性模版提取候选产品特征集，然后使用剪枝规则对该集合进行过滤，最后基于网页排序算法中HITS算法的思想对候选产品特征进行排序和筛选，得到最终的产品特征集合。实验结果表明，该方法对产品特征提取的F值达到77.3%。 2. 由于web中网页是动态变化的且不会告知搜索引擎这种变化，因此搜索
14、引擎需要不断重复访问相同的网页，以保持本地网页库的新鲜度。如果能够对网页更新的可能性进行准确的预测，网络爬虫就可以只重复访问发生变化的网页，这样既保持了网页库的新鲜度，同时也节约了网络带宽及其他宝贵的计算资源。为此，本文提出了一种基于强化学习的网页更新预测算法，该算法使用马尔科夫模型对爬虫抓取网页的网络环境进行建模，根据已抓取网页的变化情况来预测网页未来可能的变化量，并将真实的网页变化量作为反馈不断调整模型，使模型更加适应环境，从而进行更准确的预测。充分的实验结果表明，基于强化学习的方法可以达到74.76%的网页更新预测准确率，优于文献中广泛使用的泊松过程建模方法。此外，本文算法可以直接对新网
15、页进行更新预测，无需利用相同网页长时间的历史变化轨迹。3. 随着互联网的快速发展，网络舆情成为政府部门和企业以及社会大众关注的焦点，对网络舆情进行有效监管和正确引导是当前亟待解决的问题，而话题热度预测是舆情监管和引导的基础。针对现有算法无法对新话题的热度进行有效预测的缺点，本文提出了一种基于K近邻的新话题热度预测算法。该算法利用与新话题相似的历史话题的点击数时间序列来对新话题的热度进行预测。实验结果表明，在允许相对误差分别低于10%、20%和30%的情况下，算法预测的前三天点击数的平均正确率分别为47.26%、61%和67.7%，点击数变化趋势平均正确率达到73.73%，同时也说明了相似的话题在话题出现的初期具有近似的热度变化趋势。
