Search is sometimes difficult to help users find valid information

In today's rapidly expanding data and information, even search is sometimes difficult to help users find effective information. And intelligent push can enable users to master valuable information in time, so as to save people in the torrent of information.

Now relying on large data mining and cloud computing, personalized information push is not a problem. When you open a Web page and the application interface, with others open the home page and application interface results may be completely different.

Yahoo's Onepush project also targeted personalized push information. For example, Yahoo Weather Service, if Beijing is going to have a sandstorm tomorrow, will be pushed to the user today, because this is an urgent matter. If Onepush perceives that you are using the iphone and pushes this weather message, it will not be pushed on your other devices to avoid causing harassment.

It is what Yahoo is doing to make the Internet smarter and more personalized in everyone's eyes. The following is Yahoo Beijing Global Research and Development center of the general architect Ju Jinsheng to Onepush thinking and interpretation:

"Onepush" is Yahoo Beijing Global Research and Development Center (hereinafter referred to as Yahoo North Research) launched a new push technology, it will be Yahoo to improve the user experience is one of the important means. One might think: Push is already a basic technology, what else can be discussed?

But I would like to ask: Today's push technology is really valuable? Before you answer this question, let's take a look at some typical application scenarios.

Push-Send anxiety disorder

For a friend of mine, clearing the notification center has become a daily "homework" he must do. He is afraid to miss some useful notice, but the long-term use told him that valuable information is scarce, the notice center reduced to meet his digital cleanliness of the export, he frankly, clean out a variety of push notice can get a little sense of achievement.

Take a look! We have been accustomed to the push technology just created a group of "push anxiety disease" patients, which completely deviated from the original intention of the push technology. The core problem is that today's push technology is still not human, intelligent and personality. I think that the push technology itself is very valuable, but most technology companies are not making good use of them.

Six major ills of push technology

In order for the push technology to really improve the user experience, we first need to find out all the ills of the existing technology.

First, data loss. The existing push technology does not guarantee that all information can reach the user terminal.

Second, invalid push. Users receive a lot of information that they don't care about.

Third, delayed arrival. Much of the delay in pushing information has exceeded the user's tolerance limit.

Four, expired push. I often go to the United States to have a meeting, so there is a special mobile phone in the United States. When I came back, when I opened the mobile phone at home, the "good" push technology would send me all the expired information that had been stored for one months, causing my phone to become paralyzed instantly.

Five, repeat push. Many people have desktops, laptops, smartphones, tablets, and so on, and some even have two of them. No one wants to see the same information again on all these terminals.

Six, broadcast-type push. Everyone can hear the radio, but not everyone is listening.

Push technology still has so many problems, so it is very promising. With the rise of cloud computing and large data, the revolution in push technology has fertile ground, through the Yahoo Cloud computing platform, we have a large number of Yahoo global 700 million users to analyze and dig, so that we have the opportunity to develop a truly human, intelligent and personalized push technology.

To create a dedicated push technology for everyone

"Onepush" This name has three meanings, on the one hand refers to pushes the platform the unification, the Unification platform may support all terminals, on the other hand is to Yahoo all type application support, including the Web application, finally is to the application push and the application pushes the integration.

I am certainly excited to be able to lead such an important project, but there are "bad news": We have to complete the first official version within 3 months. It was a crazy decision, and I had to work on getting the whole team to change the development mindset and focus on overcoming the most valuable challenges to the user.

The development team consists of more than 20 engineers in development, experimentation, services, and products, and we still employ a relatively conventional "agile development" model, with the emphasis on fundamentally changing the way team members think. For these passionate young people, the technology is not a problem, habit is the problem-they always get used to doing it first, try again, think over and change, this practice is a bit like sculpture, take a stone, hammering, remove useless things, leaving is the work. In fact, engineering development should be more like injection molding, that is, you have already figured out what your work should look like before you do it, and do a good job.

Therefore, I ask the team members to think first, from the macro to the details, comb out the most valuable things to the user, and then concentrate firepower, a molding. This is my philosophy of development--"taking facts as the leading development".

At the beginning of June, "Onepush" was delivered on schedule and was pioneered in the use of heavy applications such as Flickr, Yahoo! Mail, Yahoo! Cricket.

The pursuit of the user experience has no end

"Onepush" research and development is a vast and long work, on the one hand, we designed for it a new logical and physical structure, rather than simply the original platform to merge and improve. On the other hand, the excavation of large data and the improvement of human-computer interaction can not be done overnight, the need for long-term accumulation and improvement, and in this process, people's behavior habits and focus of attention is also changing at all times, which requires us to have a keen insight, continue to bring users a better application experience.

The current version of "Onepush" has received quite a lot of positive feedback, and its goal is to "deliver the right amount of personalized information to the correct users in a timely and appropriate place", and is the direction of continuous improvement and effort in the future of "Onepush".