User Behaviors, Engagement, and Consequences
Loading...
Paper Number
2492
Paper Type
Completed
Description
The cold-start problem is salient in the current online two-sided markets. Theoretically, machine-generated content can serve as the information signal, thus helps to address the cold-start issue. To empirically examine whether machine-generated content helps mitigate the cold-start problem, we study the impact of machine-generated content on dataset adoption in a leading online public dataset community (Kaggle). We found that machine-generated content helps to solve the cold-start problem in the online public dataset community in two ways. First, we show that machine-generated content shortens the time needed for datasets to get their first adoption and that machine-generated content increases the dataset adoption in the initial stage and increases the dataset diffusion rate. Second, our results show that the positive effect of machine-generated content is stronger for dataset sharers who lack reputation signals. Our research opens a discussion on the role of machine-generated content in mitigating the cold-start problem in online public dataset platforms and also other two-sided markets.
Recommended Citation
HOU, JINGBO and Chen, Pei-yu, "Can Bot Improve Equity? Machine-generated Content Mitigates Cold-Start Issue" (2021). ICIS 2021 Proceedings. 24.
https://aisel.aisnet.org/icis2021/user_behaivors/user_behaivors/24
Can Bot Improve Equity? Machine-generated Content Mitigates Cold-Start Issue
The cold-start problem is salient in the current online two-sided markets. Theoretically, machine-generated content can serve as the information signal, thus helps to address the cold-start issue. To empirically examine whether machine-generated content helps mitigate the cold-start problem, we study the impact of machine-generated content on dataset adoption in a leading online public dataset community (Kaggle). We found that machine-generated content helps to solve the cold-start problem in the online public dataset community in two ways. First, we show that machine-generated content shortens the time needed for datasets to get their first adoption and that machine-generated content increases the dataset adoption in the initial stage and increases the dataset diffusion rate. Second, our results show that the positive effect of machine-generated content is stronger for dataset sharers who lack reputation signals. Our research opens a discussion on the role of machine-generated content in mitigating the cold-start problem in online public dataset platforms and also other two-sided markets.
When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.
Comments
21-UserBeh