И.С. Гудилина, Л.Б. Саратовская, Л.Ф. Спиридонова - English Reader in Computer Science (1114139), страница 6

Просмтор этого файла доступен только зарегистрированным пользователям. Но у нас супер быстрая регистрация: достаточно только электронной почты!

Текст из файла (страница 6)

If the word-processing revolution threatens to render you obsolete do not despair. If you can speak, you may yet survive, thanks to James and Janet Baker, proprietors of Dragon Systems Inc., a small software-research firm in Newton, Massachusetts.

This husband-and-wife enterprise aims to introduce the first natural language speech-recognition package for a personal computer. If they succeed, you will be able to simply dictate to your computer and watch your words appear magically on the screen.

Most speech-recognition researchers say this is at least five years away, if not 10. But according to Janet, Dragon's president, a speech-transcription program with a vocabulary of at least 10,000 words will lit the market by year's end. The entire package — desktop computer, modifications, and software — will cost $ 10,000 to $15,000. The system will be limited, requiring the speaker to pause after each word. But it will still be faster than writing in longhand.

Starting out in biophysics, Janet became interested in how biological systems process information. "While knowledge of how humans or other organisms receive or send signals is pretty low," she says, "we do not dismiss the potential of that knowledge. And we have profited by incorporating some of those ideas." For example, research on how neurons in the ear process sound signals has led to more effective speech-recognition systems.

James says that teaching a computer to take dictation is an enormous computational task. "You need knowledge of the language, of the syntactic structure of sentences, of semantics, and of the way things fit into paragraphs. We try to do all those things efficiently on a PC." Emphasis on computational efficiency sets the Bakers apart. While some private and university research teams have developed experimental speech-recognition systems beyond the capabilities of Dragon's, those systems also require hundreds of thousands of dollars' worth of hardware.

Researchers in speech recognition tend to favor one methodology or another. The Bakers refuse to commit Dragon to any one approach. "We're willing to try virtually anything." Janet says. "We are consciously integrating ideas of all kinds." This open-mindedness. she adds, has been a major factor in the company's success — it has been

profitable from the start.)

Perhaps the key to Dragon's future is the Bakers' firm belief that speech recognition can really work. "We don't want our technology sitting around as a laboratory curiosity," says Janet. "We want to get it out where people can actually use it."

NOTES:

1. to render you obsolete (v.) — оставить кого-то позади.

2. to hit (v.) — ударять, поражать, попадать в цель.

3. to dismiss (v.) — отпускать, распускать, удалять.

Unit 4

I. Read and translate the text orally at sight.

Few of us really enjoy performing software maintenance. Just as "I don't do windows" became an infamous cry of housekeepers everywhere, software maintenance tops the list of programmers' least favorite duties. But with an increasing supply of products requiring corrective changes, modifications and enhancements to keep up with user demands and tight budgets, software maintenance is not going to go away.

The statistics are staggering: 50% of MIS/DP budgets are said to be allocated to software maintenance every year. Programmers reportedly spend 50% and, in some cases, 80% of their time on ongoing maintenance. Software maintenance is said to be a multibillion-dollar industry. Industry watchers contend that more than $30 billion is spent annually worldwide on maintenance. The U.S. alone spends more than $10 billion annually.

Recent work at MIT indicates that for every dollar invested in new system development, an additional $9 must be earmarked for maintenance during the system's life cycle....

The keys to increasing productivity in almost any endeavor are the "three Ts" — techniques, tools and training. The three Ts are required for increasing programmer and manager productivity in both development and maintenance.

A collection of maintenance programming and management techniques is the essential base for increasing software maintenance productivity. The techniques can either be generic, such as problem-solving procedures, or be internal to the program or consist of external documentation. Some techniques can be specific to a particular software package or a programming language. A company should also put forth its own techniques for the effective use of software tools in a maintenance environment.

Without ongoing formal training of both programmers and managers in the use of maintenance techniques and tools, a company cannot achieve improvements in maintenance productivity. However, all the technological tools and fixes in the world won't help your organization's software if the people performing the function don't know how to use them. Believe it or not, not many MIS/DP professionals are formally trained for a job that accounts for more than one-half of their job responsibilities—that of software maintenance!

This situation must change. The mass education of MIS/DP professionals in software maintenance is in order. Education and training can tie together the materials on maintenance technologies and tools and prepare professionals for performing the vital function of maintenance.

Professionals may pursue specialized training when appropriate. Also, general maintenance training should be ongoing. Companies need to develop education and training materials-

The programming profession is very demanding but at the same time exciting and sometimes frustrating. While meeting day-to-day demands— some requiring overtime—it is very difficult to think of anything else. But MIS/DP professionals should take some time off occasionally and try to look at their professional and personal goals- Some self-analysis can be revealing. Also, set up your own productivity improvement program. Such a self-improvement program can include studying articles and reading selected professional books regularly.

NOTES:

1. enhancement (n.) — расширение, улучшение.

2. earmark (v.)—ассигновать, выделять.

3. endearor (v.) — пытаться, прилагать усилия.

Unit 5

I. Read and translate the text orally at sight.

Newest Cars Will Even "Learn" from Experience

A new generation of elevator technology is working its way into the world's office buildings, bringing with it a promise that many New Yorkers will find irresistible: the savings of dozens of wasted seconds each day.

The result is, in effect, a thinking elevator—a hybrid of programmable computer chips and high-technology sensors. Advanced versions will appear in new skyscrapers this year and next, and in many existing buildings the technologies are being grafted onto old cars and motors.

Elevator designers hope to save either passengers' waiting time or, by carrying more people with less equipment, building owners' money. Clever programming alone—apart from increases in elevator speed—can serve as many as 1,500 people an hour with elevators that formerly reached their limit at 1,000.

The smart elevator will sense not only how many people are riding but also how many are waiting at each floor. In buildings with just one or two elevators, the microprocessor has little to add, but where banks of multiple elevators handle heavy, complex traffic, the computer comes into its own. Using techniques borrowed from game theory, it makes rapid choices, sending cars to where they can do the most good, considering the system as a whole. The next leap forward will be learning ability, enabling elevators actually to anticipate riding patterns.

The idea is to stamp out the various bug-bears of elevator traffic. There is the phantom pickup, for example — the door of car number two opens to greet a rider who is no longer there, having just boarded car number one. There is car bunching —elevators all heading upward in packs around the 20th floor while impatient crowds wait at the lobby.

"In today's world an elevator systems gets overloaded and overpopulated, and the elevators tend to bunch together" even when passengers are distributed more or evenly, said William S. Lewis of Jaros, Ваши & Bolles, a consulting engineering firm in New York. Any system with many imperfectly coordinated moving parts can fall into an undesirable pattern. To overcome that. the computer has to overrule the tendency of every car to try to answer the nearest call.

In the past, when elevators were controlled by relays, simple electromagnetic switches, complex logic was impossible, and individual cars were mostly on their own. Now the computer lets elevators look at one another and adjust to changing

circumstances.

"It can say, 'Skip these calls and run express to the lobby, even though someone is going to wait longer,'" Mr. Lewis said. "That gets to be a very esoteric algorithm."

NOTES:

1. irresistible (adj.) - непреодолимый.

2. to graft (v.) — прививать, переносить.

3. to stamp out (v.)—подавлять.

4. (car) bunching (n.) — скопление, столпотворение, затор.

Unit 6

I. Read and translate the text orally at sight.

I started six years ago to look into distributed control of consumer electronics devices. It was not an academic research project studying programming languages: Doing language research was actively an antigoal. For the first three years, I worked on the language and the runtime, and everybody else in the group worked on a variety of different prototype applications, the things that were really the heart of the project. So the drive for changes came from the people who were actually using it and saying "do this do this, do this."

Probably the most important thing I learned in talking to the folks building TVs and VCRs was that their priorities were quite different from ours in the computer industry. Whereas five years ago our mantra was compatibility, the consumer electronics industry considered secure networking, portability, and cost far more important. And when compatibility did become an issue, they limited notions of compatibility to well-defined interfaces— unlike the computer industry where the most ubiquitous interface around. namely DOS, is full of secret back doors that make life extremely difficult. One interesting phenomenon that has occurred over the past five years is that consumer electronics concerns have become mainstream software concerns as the market for software in the home has grown.

The buzzwords that have been applied to Java derive directly from this context. In the consumer electronics world, you connect your VCR to a television, your telephone to a network. And the consumer electronics industry wants to make these kinds of networked appliances even more pervasive. Architecture neutrality is another issue. In the consumer electronics business, there are dozens of different CPU types and good reasons for all of them in their individual contexts. But developing software for a dozen different platforms just doesn't scale, and it was this desire for architecture neutrality that broke the C++ mold — not so much C++ the language, but the standard way people built C++ compilers.

NOTES:

1. ubiquitous (adj.) — распространенный.

2. buzzwords (n.) — термины.

3. pervasive (adj.) — всепроникающий, всеобъемлющий.

4. mold (n.) — форма, шаблон.

Unit 7

I. Read and translate the text orally at sight.

Over the past several years the University of Minnesota, in conjunction with the Institute for Mathematics and its Applications (IMA), has developed programs in industrial mathematics (both undergraduate and graduate) with strong links to local industries. The IMA has had a program of two-year industrial postdoctoral positions, with about four in residence each year. The duties of these postdocs include intemship in an industrial setting, working under a specific mentor on specific problems, as well as participation in the regular IMA program. This relationship has deepened ties between the IMA and industry, as well as brought important problems from industry into the fabric of the regular program. These have been featured in a biweekly seminar in industrial mathematics, presented largely by these postdocs and their mentors. The success of this interaction led, last year, to the creation at the University of Minnesota of the Minnesota Center for Industrial Mathematics (MCIM), which now oversees all aspects of the involvement of academic mathematics in the industrial setting at Minnesota.

One of the central concerns of the MCIM is to see this successful interaction broadened so as Avner Friedman is director of the Institute of Mathematics and its Applications (IMA) and the Minnesota Center for Industrial Mathematics (MCIM). He is also professor of mathematics at the University of Minnesota.

Fadil Santosa is associate director for the Minnesota Center of Industrial Mathematics (MCIM) and professor of mathematics at the University of Minnesota. His aim is to include many universities which have developed, or which have interest in similar programs. Thus, one of its first projects was to sponsor this workshop. It brought together over fifty mathematics faculty from such universities and industry researchers who have been working together with mathematics faculty and students. The goals of the workshop were:

1. To gain better understanding of the opportunities in industry for students who wish to pursue careers in industry.

2. To develop a document explaining to departments that wish to consider a graduate program in industrial mathematics what steps need to be taken.

The publication of the National Research Council report "Graduate Education in Science and Engineering" in the fall of 1995 recommended, among other things, the broadening of the intellectual content of the Ph.D. by introducing more interdisciplinary courses and an increase in the diversity of skills acquired during this training by adding a minor to the course of study. These objectives were also perceived In a "Workshop on graduate Student and Postdoctoral Education and Training" (www.nsf.gov/mps/workshop. htm), sponsored by the NSF Directorate for Mathematical and Physical Sciences, which also suggested increased use of off campus intemships and other real-world experiences The concept of industrial mathematics is not new.

Характеристики

Список файлов книги