Hegemony in Social Media and the effect of recommendations

Companion Proceedings of The 2019 World Wide Web Conference(2019)

引用 18|浏览24
暂无评分
摘要
As today’s media landscape is carved by social media endorsements and built on automated recommendations, both of these are often criticized for inducing vicious dynamics, such as the filter bubble effect, echo chamber, or polarization. We introduce a new model featuring a mild version of homophily and two well-known popularity dynamics. These broadly reproduce the organic activity and the algorithmic filtering, respectively, of which the latter is now commonplace within social media or other online services. Surprisingly, we show this is all that is needed to create hegemony: a single viewpoint (or side) not only receives undue attention, but it also captures all the attention given to “top trending” items.
更多
查看译文
关键词
fairness, homophily, random walks, recommendation
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要