Rebirth of the Cyber ​​Hegemony

Page 767

"Yeah, it's not like you don't know. I usually like to argue the most, so I like to watch their videos."

"That too."

Larry nodded.

But after finishing speaking, he was suddenly startled: "Professor Daim, there seems to be something wrong here?"

"what is the problem?"

"How does Huanyu Video know that you like to watch videos of bullshit?"

"This is..."

This statement also stumped Professor Daim.

He just liked to watch videos before, but he didn't pay attention to how it works.

At most, he just analyzed it. The short video mode of Huanyu Video is very good and very innovative.

Especially with the addition of music recently, the short video became popular in an instant.

But if you want to ask how Huanyu Video knows that he likes to watch the video of provocative, who can figure it out?

"Is it because I have found this type of video?"

Daim tried to analyze and said.

"That's one of the reasons."

Professor Larry nodded.

Push through keyword search, which is what many websites will do.

Especially for search engines.

Among the current search engines, the recommendation effect through keywords is very good.

However, is it just keyword push?

In Professor Larry's impression, he has never searched for words such as beauty.

But Huanyu Video is like a roundworm in his stomach, recommending a series of beauties to him from time to time.

Although he didn't like watching it just now, it was only for a while.

How could he not like looking at beautiful women? His greatest hobby in his life was looking at beautiful women.

As for why he can still become a professor at Stanford, it is because Professor Larry thinks that if he is better, he can see beautiful women better.

"Perhaps, there are others."

Leaving Professor Daim's office, Professor Larry started a market survey.

He did a survey data of 100 people at Stanford.

Although these people's current occupations are still students, the videos they produced in the Huanyu video are completely different.

Some students are more likely to watch some food videos.

Some students posted funny videos.

There are also videos of various types of cars.

There are also videos of various techniques for picking up girls.

It can almost be said that among the 100 students, the videos produced by them are completely different.

Moreover, through interviews with 100 students, it was found that these videos are actually the interests of these students.

Why does the student brush the video of eating, because he is very interested in eating.

Why did I get the car video, because this student is a car fan.

Why don't you get the same ones as them, because your interests are completely different from theirs.

Then.

Everything seemed to be figured out.

The core mode of Universal Video is not short videos, short videos + music.

Rather, they learned what users are most interested in through a series of means.

Moreover, they get rid of the push method of traditional Internet sites by continuously recommending products that users like to users.

In the end, relying on this unique push method, they successfully created Universal Video to cover everything.

"So it is, so it is."

"I finally get it, I finally get it."

After all the investigations were over, Professor Larry laughed excitedly.

It's actually not that mysterious.

The reason why Universal Video knows you so well.

It's not that he has any magic, nor does he rely on some mysterious power.

He doesn't just rely on keyword searches, he relies on all big data that is closely related to user behavior and identity.

And through the integration of these big data, and then make an extremely rigorous algorithm.

Then, in the end, you will find that any click you make on Universal Video, any page you browse, and any time you stay on a page... will become a symbol on your body, representing your interests.

However, although he looks extremely simple.However, it is such a simple thing, but he has stumped countless people around the world.

Even Professor Larry seems to understand now, but he only understood after others made it.

If this model has not been made, Professor Larry believes that he may not be able to imagine that the Internet can still play like this until he dies.

Chapter 868: One step behind, every step behind

"Page, let me tell you, I finally figured out the core model of Universal Video."

Excited, Professor Larry called Google Page.

The phone call ended, and like Professor Larry, Page ran away in excitement.

"Brin, Brin, Professor Larry told me the core model of Universal Video."

"Really, Paige, what is it, don't get excited, just speak slowly."

Immediately, Page, like Brin, screamed completely in the office.

"so amazing."

"I didn't expect Universal Video to think of such a model."

"Brin, we didn't deserve to lose."

Paige said with great emotion.

"No matter what, we have to make our core algorithm."

"Yes, we are going to push this one now."

The two founders of Google started out as search engines.

If it is a little difficult for others to understand this part, but they are just a little bit, and they are very transparent all at once.

There is no doubt that this big data algorithm is much more powerful than relying solely on search engine keywords.

You know, search engine keyword push is passive.

Because he needs the user to search before the search engine can know what you want.

But if the user does not search for keywords, do you know what the user wants to do?

However, Huanyu Video has taken a different approach, conducting precise analysis from the user's behavior habits and identity characteristics.

The length of time you watch the video, your mouse clicks, your likes... can all become data representing your preferences.

Combining these data together, you can make accurate recommendations for any user.

That's why Huanyu's video push is what users like.

Of course, there is also a problem of Internet logic.

In the past, everyone engaged in the Internet to let users watch whatever they pushed.

But in the Internet age with more and more information, users may not like anything you push.

What they like is what they like.

This sentence seems to be a bit convoluted, but with a change in thinking, the operation mode of the entire website will completely change.

However, when the two were excitedly discussing that Google would also come up with this big data algorithm...

Suddenly, Paige completely lost interest, and sat down on the office chair decadently.

"Peppa, what's wrong with you?"

"Peppa?"

Brin patted Page on the shoulder.

"Don't shoot, I'm fine."

Paige sighed.

"I was scared to death. I was going to call the doctor just now, why?"

"I was thinking, is it still necessary for us to come up with big data algorithms?"

"Of course it is necessary."

Brin said excitedly: "As long as we also come up with such a big data algorithm, then we can surpass Huanyu Video."

"Look, beyond."

Paige smiled wryly: "It's only been a while, and Huanyu Video has become the target we want to surpass."

Obviously their Google is the leader in the video business.

This also made Brin a little embarrassed, but he quickly adjusted: "Page, it's okay, temporary victory or defeat is nothing, as long as we work harder, we may not be able to surpass them."

"So how long will it take us to come up with such an algorithm?"

"This one……"

Brin calculated in his head.

Although such an algorithm is simple to calculate, it is definitely not so easy to come up with it.

If they just randomly come up with a big data algorithm, they can do it in a day.

However, the most important thing about big data algorithms is not the algorithm itself, but the analysis of the proportion of various user behaviors.

Because users have various behaviors when operating on the computer, how much does click account for in the entire algorithm system?

How many likes are there?

How much time do you spend viewing videos?

How should the proportion of watching for 3 seconds and watching for 5 seconds be divided?

This is not something that can be figured out just by thinking about it. He still has to conduct experiments in reality.

This experiment alone will take at least a few months.

At this thought, Brin also frowned.

At present, Universal Video is already ahead of their Google. Of course, this is not just a lead in IP data.

The most important thing is that the entire core algorithm is leading.

This lead is not just a lead in time.

This lead is still the logical lead of the entire algorithm.

The Internet is always one step first, step by step first.

One step behind, you will fall behind others step by step.

On such a core algorithm, even if it is developed by Google, their algorithm is far inferior to that of Universal Video.