基于大数据的新闻内容生产效能评估——以上海某媒体为例

An Evaluation of the Production Efficiency of News Content Based on the Big Data:A Case Study of a Media in Shanghai

在新媒体数据异构化管理的背景下,如何有效地进行新闻内容生产的绩效评估并提出针对性的方案,以提升媒体人的内容生产力成为融合性新闻机构的一大挑战。论题使用上海某新闻媒体约三年间的全部真实产出数据,通过深度神经网络对其中的规律进行了学习,从而得到了融媒体内容生产的效能预测模型,拟合出了内容产出与预期指标规划之间的关系,据此对实验对象从数据特性层面到社会规律层面进行分析。经过测算,这一模型在该数据集上可达到平均90%以上的准确率和召回率。根据模型,该媒体最优的预期新闻产量是在预期总工时达到42012小时,即126名在职新闻工作者大约每月工作时长约333小时的时候,实际新闻产量可达到47106小时的效果。这一研究可以依据现实情况,对媒体的产出情况进行最优规划。既为当下媒体融合中不断收集完善的大数据指出了利用之道,也可对其他知识性生产领域的效能评估提供借鉴参考。

Under the background of the isomerization management of new media data,a big challenge for the integrated news organizations is the way of effectively evaluating the performance of news content production and proposing the targeted plans so as to improve the content productivity of media people.With all the real output data of a news media in Shanghai for about three years,this paper studies the rules through the deep neural net.work.Then an efficiency prediction model of the content production in the integrated media is obtained to match the relationship between the content output and the expected index planning,which is utilized to analyze the exper.imental objects from the level of data characteristics to that of social law.After the mock exam,the model can achieve an average accuracy rate and recall rate of 90%or more.According to the model,the optimally expected news output of the media can be achieved when the total of expected hours reaches 42,012 hours,or in another way,126 journalists work for about 333 hours a month and,the actual news output can reach 47,106 hours.This study can help realize the optimal planning for the media output according to the real situations.It not only points out the way to use the big data which are constantly collected and improved in the current media convergence,but also provides reference for the performance evaluation of other knowledge-based production fields.