Using social-sharing mechanisms to refer to content

We've talked a lot about personalized reading, such as "Today's headlines" based on social-networking data mining, such as "finger-reading", which uses social-sharing mechanisms to recommend content, and so on, and now there's one more player in the field.

The company, called "The Fun", is a mobile app that provides personalized reading, and they choose a new perspective to interpret "personalized reading".

"We use Google mode to refactor personalized content." Johnny, co-founder of Li Zhong, explained to me: "Just as Google recommends Web pages by computing the degree of association between pages and pages on the Internet, the blog has a system called" keyword graph "that calculates the strength and weakness of keywords and keywords in the article. Through these strong and weak relations to the relevant content of the order, and according to the ordering of the reader to recommend the more relevant content. Meanwhile, due to the characteristics of time, the key words can be changed according to the content of the network and the reading behavior of the users, so as to personalize each reader's reading experience constantly. ”

In fact, I did not understand after listening, Johnny gave me a practical example: "For example, you pay attention to Alibaba, usually, you will be interested in Alibaba, Jack Ma, so ' ma ' this key word for ' Alibaba ' is ' Strong contact, so when you use the blog to read Alibaba related content, and Ma Yun related content will also be recommended. And this is not the same, such as Alibaba recently invested in Sina Weibo, so ' Sina Weibo ' in the near future also belongs to Alibaba strong key words. Therefore, the long-term use of the blog, users will be more accurate reading content, at the same time, there will be no narrowing of the problem of information. ”

The method used by the volunteers would be quite high for early data, otherwise the keywords would not be effective. At present, the blog has been through in-depth cooperation with Sina Weibo, by pulling the user micro-blog content, to obtain a number of users interested in keywords, these keywords as a blueprint for users to recommend content. And enjoy a year-long time, in 600 or so users in the beta, during the repeated communication with the user, the correction algorithm to achieve relatively accurate.

To say that App,app design is very good, smooth operation. In the initial use, in addition to the use of micro-blog login, you also need to choose their areas of interest, as well as a part of the relevant areas of keywords. In particular, each piece of content below "share", "point praise" "comment" Three keys, and "point praise" and "comment" articles will not sync to their own flow of information, only "sharing" can.

At the same time, the blog also provides a ' circle of friends ' to recommend other users who have a congenial interest in the user's reading behavior to help filter out targeted content. "Of course, if you don't like it, you can just focus on your own personalized magazine experience, regardless of what these friends share, which is only a supplement." Johnny stressed.

At present, the team has only five members. The two founders were technical, Johnny at the University of Southern California, majored in software and hardware, graduated with Master and PhD, and successfully started a business in the United States, creating a website for home appliances, which is still working. Another founder, Allen and Johnny, was a college alumnus and worked as a senior software engineer at Qualcomm. Other members include an iOS development engineer, a graphic artist and a backstage engineer, all five people, as shown in the figure below.

Johnny told me in an interview that the team is now very understaffed and needs Android and iOS development engineers. Currently enjoy in Shenzhen office, interested in children's shoes can be seen in detail.