How to use artificial intelligence to successfully shoe the user

The news site Variety recently wrote about how media companies can use artificial intelligence to connect with consumers and recommend more personalized content.

The following is the original content:

How to use artificial intelligence to successfully shoe the user

Killer computers, robotic riots - Hollywood has long since assumed the negative effects of artificial intelligence. Even outside the silver screen, artificial intelligence has gradually become an important part of the media industry - but it has not yet reached the advanced level described in the films "She" and "Mechanical Ji".

For example: You may have had contact with artificial intelligence today, but you don't know it. It pushes notifications to your phone, and your favorite website sends you emails, and it also recommends content for you as you watch the video – all with machine learning. They use massive amounts of data to guess what media content you might like.

YouTube’s content recommendation system is one of them. In the early days, the video-sharing site just used a model called collaborative filtering to recommend content to users. If 10 people have selected the same 5 videos, the site will expect the next person to watch 4 of them will also like the 5th paragraph.

Over time, YouTube has refined the recommendation model and adopted a more sophisticated forecasting model, but still does not adequately cover the massive video on its website.

"YouTube is so diverse," said Christos Goodrow, vice president of YouTube search and discovery technology. "We think everyone on the planet can find 100 hours of video on YouTube. ""

But the key is how to accurately find this 100-hour video. For this reason, YouTube is actively embracing deep learning techniques based on neural network technology that can use algorithms to simulate how the human brain works. These algorithms can be used to train smart decisions and recommend the right video for the user.

Goodro developed a similar neural network for large enterprise customers 20 years ago. At that time, the technology was still somewhat purposeless. “Now, they suddenly become very effective.”

He explained that the key difference lies in the data. YouTube's algorithms are based on hundreds of billions of viewing behavior data points—which were simply impossible to achieve a few years ago.

Google’s parent company, Google (Weibo), has seen artificial intelligence as a key component of its future business, and Google CEO Sundar Pichai recently told reporters at a press conference that the artificial intelligence revolution and the Internet or intelligence The invention of mobile phones has the same profound significance.

“We are at a groundbreaking moment in the computing industry.” Pigue believes that “the world is evolving from the era of mobile first to the era of artificial intelligence.”

Alex Holub, CEO of San Francisco's artificial intelligence startup Vidora, agrees. "In the next three to five years, every company will deploy an artificial intelligence engine at its core," he said.

Vidora is helping media companies such as Vudu Video Services, which is owned by News Corp., Yahoo Japan and Wal-Mart, to promote artificial intelligence projects. The technology of this startup can help these companies customize the communication with users to reduce the rate of churn, either through websites and applications, or through push notifications on mobile devices.

Vdora's artificial intelligence-based system not only learns which videos should be recommended, but also how to promote each video and even find the best price and value proposition for each customer. “Everyone can get their own experience,” Horub said.

To achieve this goal, it is also necessary to use the huge data to help the system understand user habits, and then train the artificial intelligence engine. "In the online world, you can get exclusive data for each user, which was not possible before." Horub said.

Horub insists that over time, these systems will gradually improve and perform complex tasks. For example, a video subscription service can use the artificial intelligence system to determine the copyright of which movies should be purchased in the near future.

Goodro also hopes to use artificial intelligence to help YouTube meet the major challenges of the future: understanding customer satisfaction. This service already knows which videos users like to watch, but doesn't really understand what these users like. The site is trying to get feedback on each video through a survey. After a while, they may be able to use this data to train artificial intelligence algorithms to identify subtle signals that reflect user satisfaction.

In other words: artificial intelligence does not become a killer robot, but can easily become a killer content recommendation engine.

DC Charger

Professional DC Charging Station manufacturer is located in China, including DC Ev Charger,DC Charging Pile,DC Fast Charger, etc.

DC Charging Station,DC Ev Charger,DC Charging Pile,DC Fast Charging

Shenzhen Hongjiali New Energy Co., Ltd. , https://www.hjlcharger.com