Communication of Smart Media by Peng Duan & Lei Zhang & Kai Song & Xiao Han

Author:Peng Duan & Lei Zhang & Kai Song & Xiao Han
Language: eng
Format: epub
ISBN: 9789811594649
Publisher: Springer Singapore

4.2.3 Smart News Distribution

The focus of this section is on the application of intelligent technology in news distribution. Of all the parts of journalism, intelligent technology is best applied in news distribution. It mainly uses algorithms to recommend to users’ news products in which they may be interested, so it is called algorithm news by some people as well. In 2016, China Internet Network Information Center pointed out in its 38th “Statistical Report on the Status of Development” that algorithm distribution based on user interests was gradually becoming the main way of distribution for online news.

Personalized recommendations are the most important principle of algorithm news, which fundamentally solves the problem of information matching through machine algorithms. Traditional news distribution is completely tied to news production and is done by professionals. The rise of algorithms makes it possible to record and store data of users’ different behaviours in different situations, so as to realize data identification and screening and match the data with user profile and user needs. Based on continuously optimized algorithms, it sets personalized agendas for different users. For example, Jinri Toutiao’s news client uses the advertising slogan of “Headlines You Care About” and mainly promotes algorithm recommendation mechanism. It builds an intelligent news platform, and by analysing the user profile, scenes and article features, it recommends different personalized news to different users. How does it achieve that? There are three supporting factors here. The first one is the user profile. It learns users’ characteristics and needs by capturing user information. The second is its mass content. Using the content posted by accounts at Jinri Toutiao and content captured by crawlers, it builds a rich content of storage warehouse. The third is algorithm recommendation. Based on user characteristics, information entered and history, it quickly searches for filters and recommends relevant content.

Here the precise determination of user profile is a key point. In fact, it is e-commerce businessmen rather than Internet news applications that do best in this. Using captured information like labels, purchase history, keywords searched, people followed, friend categories, Microblog content, Comments, reposting preferences, locations, model and service time of mobile phones, and so on, e-commerce businessmen can recommend to us products we are most interested in.

What contributions do we users make to realizing personalized recommendations? First of all, we allow the aggregation services of Jinri Toutiao to access all kinds of information. We let it read our location, so it can recommend local news to us. We build a friend list. It will make recommendations based on our friends’ browsing history. Secondly, we enter all kinds of information by ourselves. For example, we search for some keywords, and it gets to know our interest and needs. Thirdly, we actively subscribe to some accounts on Jinri Toutiao. It is equivalent to a personal customization, providing the basis for the aggregation of the algorithm. Finally, when we constantly browse news, the algorithm gets to know us better. According to the Control Theory, each browse of the user counts as an entry of feedback to the algorithm.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.