My 1999

Chapter 689 Machine Translation

After chatting about this topic for a while, Xu Liang left.

Jiang Xiaoyang didn't live at home either, so he went with him, making Liu Jihong complain.

Most of them have a husband and forget about their parents.

Returning to the top floor of Hanhua Center, the two of them had the 2,000-square-meter house to themselves. After taking a shower, Jiang Xiaoyang went to bed early.

She didn't stay up late before becoming pregnant, but went to bed earlier after becoming pregnant.

I usually go to bed at 9:30 and fall asleep at 10:00.

Get up around six in the morning.

Take another half-hour break at noon.

Have lots of energy for the day.

Xu Liang didn't sleep that much. Not only was he in good health after his rebirth, he recovered quickly and was full of energy.

Four or five hours of sleep at night and a short nap at noon are enough for one day.

The summary is that both red and blue are high, and recovery is extremely fast.

I processed some work emails and after early morning, I took a quick shower before going to bed.

The next day Jiang Xiaoyang went to work as usual, and he went to Hongmeng headquarters.

"Mr. Xu."

"Lao Xie, give me a copy of the satellite data you prepared last time."

Xie Wen took out a piece of information from his safe, Xu Liang flipped through it briefly and handed it to Lu Hui.

"You personally take it to the city hall and hand it over to Liu Shucheng."

Lu Hui nodded and left.

She had been with Xu Liang several times and got to know him.

"Did the superiors promise to develop satellites for us?" Xie Wen asked.

"Not yet. I just sent out the information to inquire about it. I'm not sure whether it will work out in the end."

"Hopefully, if we can succeed, we will be able to contribute to the domestic aerospace industry."

The satellites they need are worth nearly 100 million US dollars each, and the profits are huge.

Xu Liang nodded.

"Is the open beta of "World of Warcraft" going well?"

"It went very smoothly, with only a few minor issues that needed to be corrected."

"When will it be released?"

"The North American company reported that it will most likely be launched globally in November."

Xu Liang nodded, "Keep an eye on me and don't make any mistakes."

He has been coveting this classic game masterpiece with over 10 million peak payers and over 10 billion US dollars in total revenue.

"OK."

Xie Wen agreed.

"By the way, Mr. Xu, Internet Technology is about to go public. Are you going?"

Hongmeng Network Technology Co., Ltd.’s road show has ended, and Xiangjiang’s listing time is scheduled for September 15.

Listing number 0778.

The total share capital is 1 billion shares, and 200 million shares are placed, of which 120 million new shares are sold and 80 million shares are cashed out by investors.

The price per share is HK$62, and the valuation multiple is 20 times.

The main business of Wangke is Internet cafes, with retail and advertising included.

The cost is very high, and it does not have high growth potential like the Internet, and competition is fierce, so the valuation multiples are inevitably affected.

After several months of road shows, the Internet Technology Company's subscription has exceeded HK$54.9 billion, and you can win 1 share for 4.4 shares, which is very good.

"When Hongmeng's first company goes public, of course I, as a parent, have to be there." Xu Liang said with a smile.

Xie Wen smiled and agreed, "Mr. Xu, Google will go public tomorrow."

Xu Liang responded.

Although he has changed a lot of the original history, this little cutie still has its strong inertia.

Google's time to market remains unchanged.

August 19, 2004, that is tomorrow.

Google rings the bell on Nasdaq.

However, a large piece of the foreign market was taken away by Bing.

Google's valuation in this life has been affected, with only 14 billion US dollars, almost half less than the 23 billion in the previous life.

“What’s going on with our lawsuit against Google?”

"Still fighting. Google is like a dog-skin plaster that can't be shaken off no matter how hard you shake it off." Xie Wen said helplessly.

"The courts in the United States are so pissed off that there won't be any results in two or three years."

If you lose the case in the first instance, you can still appeal, from the intermediate court to the Supreme Court, plus the process of producing evidence in the middle. It's like an old woman's foot wrap, smelly and long.

"Fortunately, we didn't suffer any loss." Xie Wen said with a smile.

"Yahoo's Yang Zhiyuan doesn't think so."

The two looked at each other and smiled.

The lawsuit between Hongmeng and Google became known to the world due to Google's deliberate publicity, which greatly enhanced Google's brand awareness.

In the same way, Hongmeng has become well-known in the English-speaking market, at least its popularity has greatly increased.

Under the squeeze of both parties, other search engines around the world either went bankrupt or were acquired.

Yahoo search originally accounted for 45% of the English search market, but now only about 30% remains.

Microsoft's MSN is also in dismal business. Even with its powerful system software and forced installation, its market share in the English market is less than 7%, which is still lower than Hongmeng Bing, which has not made much effort in the English market.

The only one who can still maintain a piece of pure land is Mao Xiong’s Yandex.

The competition between the two sides perfectly illustrates what it means to fight between the boss and the second child, but it is the third child who dies.

Xu Liang looked at the time.

"It's noon, let's go and have dinner together."

Xie Wen nodded and followed.

The two of them didn't go out either. Hongmeng's restaurant tasted good and had a rich variety, which was much better than ordinary buffet restaurants.

After finishing lunch and returning to the office to rest briefly, Xu Liang came to Hongmeng Research Institute.

Hongmeng Research Institute occupies half of the floors of Wangjing Building, which has now been renamed ‘Hongmeng Center’. The number of people has exceeded 1,300, including not only top domestic talents, but also Chinese elites hired with high salaries from abroad.

Hongmeng spends nearly 4 billion Chinese coins on Hongmeng Research Institute every year.

There are currently three main departments.

One is the ‘Hongmeng Mobile System’.

The second is big data and cloud computing developed by Pangu Company.

The third is ‘artificial intelligence’.

In addition to these three major research directions, there are also some independent R\u0026D departments, such as programming languages, office software, etc.

Each of the three major departments has different research and development directions.

For example, under the "artificial intelligence" department, there are machine translation, intelligent speech, image recognition, object recognition, and more advanced virtual reality.

Big data and cloud computing include databases, data algorithms, cloud platforms, etc.

There are even more Hongmeng phones.

"Dr. Tang, I saw in your work report that there has been a big breakthrough in machine translation?"

Tang Xiaoou smiled and nodded.

"Through everyone's efforts, in terms of machine translation, we have reduced the error rate by half and doubled the accuracy."

"Because of big data?"

"Yes. Because of the massive data provided by Hongmeng Bing, Yihaodian, etc., we adopted a six-element data model. In contrast, other research teams that lacked data support could only use a three-element model.

A good three-gram model can accurately construct the collocation between phrases and simple sentence components in English sentences, while the six-gram model can construct the collocation between the entire clause and complex sentence components.

It is equivalent to directly translating these fragments from one language to another, and the accuracy is greatly improved. "

Looking at the excited Tang Xiaoou, Xu Liang couldn't help but said.

"Dr. Tang, what is the difference between the six-element model and the three-element model?"

"To put it simply, the N-gram model considers the associations before and after N words, the six-gram model considers 6 words, and the currently commonly used ternary model only considers 3 words."

“In the field of artificial intelligence research, are many six-element models used?”

"The six-element model is not a new concept, but without sufficient data support, the accuracy of the five-element or six-element model is very poor and will not be of any help to translation."

Xu Liang understood.

"Can I understand it this way? If there is enough data, will there be a seven-element model or even a nine-element model in the future?"

"Theoretically yes, but currently we are not able to provide such a huge amount of data."

Xu Liang nodded. The amount of data is not enough now, but when we enter the mobile data era, the amount of data will increase geometrically.

"Mr. Xu, I hope to represent the company in the machine translation evaluation held by the National Institute of Standards and Technology in the United States next February."

Xu Liang frowned slightly, "Why do you want to participate? It has international influence?"

"Yes. NIST is the Everest in the field of machine translation. Whether you are a university or a company, as long as you have some research on machine translation, you can participate.

The University of Southern California, IBM Watson Labs, University of Maryland, Johns Hopkins University, and Germany's Asia Deep Institute of Technology, all of which are institutions with outstanding strength and reputation in the field of machine translation, will participate.

If you can get a ranking, it will not only greatly enhance the company's reputation, but also increase its attractiveness to senior talents. "

Xu Liang nodded slightly.

Although Hongmeng is developing well, it is still not attractive enough to top talents.

For example, in front of you is Tang Xiaoou, a Ph.D. from MIT who will be a technical expert in founding SenseTime Technology in the future. Hongmeng invited him many times and did not hesitate to pay a high salary. In addition, the other party really wanted to return to China, so he finally started from the U.S. undersea robot experiment. The room was dug out.

Originally, they wanted to poach Kai-Fu Lee, but he is now the global vice president of Microsoft and looks down on the position of head of a research laboratory under Hongmeng.

However, as Hongmeng continues to develop and agree to give options, the other party has relaxed.

If Hongmeng can make greater achievements in the field of technology, I believe it will be easier to talk about it.

"Mr. Xu, we do not receive R\u0026D funds from the US government, we have no obligation to disclose research details, and we will not leak the company's technology." Tang Xiaoou said quickly.

Xu Liang nodded.

"I agreed to this, but there is one thing. You can't go to this meeting."

"Why me……."

Xu Liang interrupted him and said seriously.

"Dr. Tang, you are our company's top scientific research talent, and you are developing the most cutting-edge field of artificial intelligence, so I will do my best to ensure your personal safety."

Although he was touched by the other party's concern, Tang Xiaoou couldn't help but said.

"Mr. Xu, are you making too much of a fuss?"

"Don't be afraid of ten thousand, just be afraid of the worst."

Xu Liang handed over a stack of information.

Tang Xiaoou turned over the pages with a look on his face several times.

The above are all news about the accidental death of some top scientists.

There are furry bears, there are Chinas, and there are even Frances and Germanys.

Regardless of whether these deaths were caused by Uncle Sam's handiwork, they did shock Tang Xiaoou greatly.

His attitude also softened.

After all, everyone is afraid of death.

Tap the screen to use advanced tools Tip: You can use left and right keyboard keys to browse between chapters.

You'll Also Like