Paper ID
2003
Paper Type
short
Description
Attention is the currency that facilitates transactions between producers (authors) and consumers (readers) on the Internet. In this sense, the distribution of attention corresponds to the distribution of income, raising concerns about attention inequality that can have negative consequences similar to the ones that are associated with income inequality. To the extent, social media plays an important role in shaping attention, and platforms have implemented algorithms to help users combat information overload, it is an open question how such algorithms would influence the attention inequality on social media. We use a natural experiment on Sina Weibo, and leverage our unique dataset that includes 4,760 newly registered users, 2.7 million followers and followees to answer this question. Our findings suggest the algorithmic filtering (1) suppresses fan attraction for unpopular users but helps to attract more interactions received by their posts at the user level, (2) helps to attract more discussion for unpopular topics at the topic level.
Recommended Citation
Li, Kayla Guangrui; Mithas, Sunil; Zhang, Zhixing; and Tam, Kar Yan, "How does Algorithmic Filtering Influence Attention Inequality on Social Media?" (2019). ICIS 2019 Proceedings. 8.
https://aisel.aisnet.org/icis2019/crowds_social/crowds_social/8
How does Algorithmic Filtering Influence Attention Inequality on Social Media?
Attention is the currency that facilitates transactions between producers (authors) and consumers (readers) on the Internet. In this sense, the distribution of attention corresponds to the distribution of income, raising concerns about attention inequality that can have negative consequences similar to the ones that are associated with income inequality. To the extent, social media plays an important role in shaping attention, and platforms have implemented algorithms to help users combat information overload, it is an open question how such algorithms would influence the attention inequality on social media. We use a natural experiment on Sina Weibo, and leverage our unique dataset that includes 4,760 newly registered users, 2.7 million followers and followees to answer this question. Our findings suggest the algorithmic filtering (1) suppresses fan attraction for unpopular users but helps to attract more interactions received by their posts at the user level, (2) helps to attract more discussion for unpopular topics at the topic level.