News application software-based user interest label construction method and related equipment
The invention relates to a user interest portrait technology in the field of big data analysis, in particular to a news application software-based user interest label construction method and related equipment. The method comprises the steps of obtaining news information browsed and commented by a us...
Saved in:
Main Authors | , , , |
---|---|
Format | Patent |
Language | Chinese English |
Published |
15.10.2019
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | The invention relates to a user interest portrait technology in the field of big data analysis, in particular to a news application software-based user interest label construction method and related equipment. The method comprises the steps of obtaining news information browsed and commented by a user on multiple types of news application software; calling a word segmentation system, constructinga keyword library, and calculating keyword weights; calculating an interest score according to news information browsed and commented by the user; and comparing the interest scores with a preset scorethreshold, screening out the interest scores higher than the score threshold, and updating the corresponding interest tags to a user tag library. According to the method, the browsing habit of the user is quantitatively analyzed, user behaviors are conveniently analyzed, and the accuracy of recommending content to the user is improved.
本发明涉及大数据分析领域的用户兴趣画像技术,具体涉及基于新闻应用软件的用户兴趣标签构建方法及相关设备,所述方法包括:获取用户在多类新闻应用软件上浏览、评论的新闻信息;调用分词 |
---|---|
Bibliography: | Application Number: CN201910245276 |