用户文章影响力分析数据
收藏资源简介：
各用户在龙游通app上发布文章，内容辐射全球各大热点事件，对每个文章的影响力进行分析，根据影响力的不同，可以帮助把握当下热点，了解用户感兴趣的方向，对文章作者进行创作引导，对影响力大的文章及热点事件进行首页推送，增加流量。1、数据采集：对平台上发布的文章进行后续数据的跟踪收集，用户在龙游通app进行操作时，会调用java服务端相应操作的接口，比如用户打开一篇文章，就会调用文章浏览接口，通过浏览接口中的代码逻辑将本次浏览记录存储到mysql数据库中，点赞、评论、收藏、转发、当天文章总数等数据也都是采用的这种方式，2、数据处理：对收集到的数据进行去重、降噪、填补缺失值等处理，确保数据的真实性，后台大数据处理系统通过不断升级、学习，对文章影响力的判断方法也会不断更新，影响力=(评论数*0.4+收藏*0.3+点赞*0.2+浏览数*0.1+转发*0.3)÷当天文章总数，将文章影响力作为判断文章、作者及该事件受关注程度的重要指标，可以了解当下热点事件、为用户创作方向指导。
Users publish articles on the Longyoutong mobile application, with content covering major global hot events. We conduct influence analysis for each article. Such analysis helps grasp current hot topics, identify user-preferred directions, guide article authors' creation, and push high-influence articles and hot events to the homepage to boost traffic. 1. Data Collection: Track and collect follow-up data of articles published on the platform. When users perform operations on the Longyoutong app, corresponding interfaces of the Java backend server are invoked. For example, when a user opens an article, the article browsing interface is triggered, and the logic within this interface stores the current browsing record into a MySQL database. Data including likes, comments, collections, reposts, and the total number of articles published on the current day are all collected via this approach. 2. Data Processing: The collected data is processed with operations such as deduplication, noise reduction, and missing value imputation to ensure data authenticity. The backend big data processing system undergoes continuous upgrades and learning, leading to iterative updates of the article influence judgment method. The calculation formula for article influence is: Influence = (Number of Comments × 0.4 + Number of Collections × 0.3 + Number of Likes × 0.2 + Number of Views × 0.1 + Number of Reposts × 0.3) ÷ Total Number of Articles on the Current Day. Article influence is taken as a core indicator to evaluate the attention level of articles, authors and related events, which enables us to understand current hot events and provide guidance for authors' creation directions.
