BEIJING, July 16, 2021 / PRNewswire / – Recently, 4Paradigm officially signed a contract with People’s Daily to jointly create consumer algorithms for new media. This will help ensure the precise mapping of massive content with the individual needs of users, to achieve the delivery of high-quality content in mainstream media and promote the transformation and innovation of the media industry in the age of AI.
Dai Wenyuan, president of 4Paradigm, said the media industry is not just about flow, but also needs to radiate positive energy. Thus, we need to modify the characteristics of the past recommendation algorithm which only optimizes clicks and optimizes user time, and adds elements of values to the algorithm. We are not only excited to have the opportunity with People’s Daily to explore and meet individual needs, but we also reflect the value judgments of traditional algorithms. While ensuring a precise match of content and user needs, a fair balance is struck between individual needs and group values.
Ding Wei, director of the People’s Daily New Media Center, said algorithms predominate in the smart age. From a thousand people see the same thing to a thousand people see their single thing interests, algorithms reconstruct the logic and the rules of information dissemination. We have cooperated with companies such as 4Paradigm to launch mainstream algorithms in People’s Daily client version 7.0 to promote the People’s Daily client’s strategic transformation from traditional media to smart media.
He highlighted the three characteristics of traditional algorithms. The first is more quality content. The creators of the People’s Daily platform are analyzed by uploading users to check the quality of the source content. In terms of content classification, it relies on the new media team and the People’s Daily review team to classify and identify content and set up a quality assessment system. Meanwhile, with the help of artificial intelligence technologies such as semantic scene recognition and intelligent deep learning technology, issues such as content redesign are solved. The second is to better understand your recommendations, perform multidimensional feature descriptions, and get an efficient and accurate match between massive content and individual user needs. Traditional algorithms comprehensively characterize user interests through analysis of long-term and short-term changes in user behavior, and dynamically characterize current user interests and preferences through real-time phenomena estimation. At the same time, with the ability to enrich the user’s offline mining interests label, meets the needs of diverse and personalized users. The third is a richer and more open information environment. Consumer algorithms can provide users with a cross-collar knowledge system and break down barriers in information cocoon rooms. Through the establishment of a user knowledge system in massive information, the integration of user behavior and semantic recognition, and the exploration of causal relationships, so that the machine can train a stronger reasoning capacity, in order to achieve the recommended content, from the point -Extension point by point to the inter-domain presentation.
The “first recommendation platform” of 4Paradigm thanks to the real-time and large-scale operation of several systems such as the quality evaluation system, the two-way interactive recommendation system of the user platform, the system text analysis and user portrait system. A special recommendation system was built from 0 to 1 for People’s Daily, and it was officially launched on the People’s Daily information client. “First Recommendation” is a large-scale machine learning-based recommendation system service platform produced by 4Paradigm, which aims to lower barriers for media to adopt new technologies. Currently, the First Recommendation Platform has conducted extensive cooperation with thousands of media and content platforms.
Artificial intelligence is rewriting the media format, the form and mode of communication of information products are redefined. Content distribution affects the distribution of media traffic and benefits, and personalized recommendations have become the backbone of major media. The innovative exploration of 4Paradigm and People’s Daily in the field of algorithms further increases the value of content delivery.
SOURCE 4 Paradigm