量化和预测演艺界的成功
Quantifying and predicting success in show business
摘要:
最近对成功科学的研究表明,科学家或艺术家最具影响力的作品在个人职业生涯中随机而均匀地发生。然而,在某些艺术活动中,比如在电影和电视中表演,有一份工作可能是最重要的成就:成功只是谋生。通过分析与电影和电视相关的大型在线信息数据库,我们能够研究那些在娱乐行业工作的人的成功。我们首先支持我们最初的说法,发现三分之二的演员是“一击成名”。此外,我们发现,在协议与以前的作品,活动聚集在热条纹,和个人活跃的职业生涯的百分比是不可预测的。然而,我们也发现,在演艺界的生产力有一系列的显着特点,这是可以预见的。我们揭示了一个富人更富的机制的存在下分配的工作,与齐普夫法律出现的总生产率。我们发现,生产力往往是在职业生涯的开始最高的,而且“奇迹之年”的位置-演员最多产的一年-确实可以预测。基于这些程式化的签名,我们开发了一种机器学习方法,该方法可以以高达85%的准确率预测演员的奇迹之年是否已经过去,或者更好的日子是否还会到来。最后,我们的分析是分别对演员和女演员进行的,我们揭示了这两个群体在不同指标上的可测量和统计上的显著差异,从而为演艺界的性别偏见提供了令人信服的证据。
英文摘要:
Recent studies in the science of success have shown that the highest-impact works of scientists or artists happen randomly and uniformly over the individual's career. Yet in certain artistic endeavours, such as acting in films and TV, having a job is perhaps the most important achievement: success is simply making a living. By analysing a large online database of information related to films and television we are able to study the success of those working in the entertainment industry. We first support our initial claim, finding that two in three actors are "one-hit wonders". In addition we find that, in agreement with previous works, activity is clustered in hot streaks, and the percentage of careers where individuals are active is unpredictable. However, we also discover that productivity in show business has a range of distinctive features, which are predictable. We unveil the presence of a rich-get-richer mechanism underlying the assignment of jobs, with a Zipf law emerging for total productivity. We find that productivity tends to be highest at the beginning of a career and that the location of the "annus mirabilis" -- the most productive year of an actor -- can indeed be predicted. Based on these stylized signatures we then develop a machine learning method which predicts, with up to 85% accuracy, whether the annus mirabilis of an actor has yet passed or if better days are still to come. Finally, our analysis is performed on both actors and actresses separately, and we reveal measurable and statistically significant differences between these two groups across different metrics, thereby providing compelling evidence of gender bias in show business.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。