Abstract
Cryptocurrency, which relies on decentralized computer software programs to manage monetary transactions, has been associated strongly with illegal money laundering on the dark web. Understanding the social networks of the cryptocurrency software development communities can provide useful investigative leads and business intelligence. However, research in modeling software social networks (SSNs) of cryptocurrency software development is not widely available. This research developed three temporal network models to predict and to simulate the activities of cryptocurrency SSNs over time. The models capture the activity history of the SSNs and dynamically combine the pricing information of the related cryptocurrencies to enhance the accuracy of prediction and fidelity of simulation. We used the models to predict and simulate the SSNs of 83,536 repositories that are related to the three cryptocurrencies: Bitcoin, Monero, and Ethereum. Experimental results show that the model (SEAM) that captures recency / primacy effects of human cognitive processing outperformed other models in metrics pertaining to influence, popularity, and engagement. The research should enhance understanding of online behavior of the SSN community and identify relationship between SSNs and cryptocurrency market.
Recommended Citation
Chung, Wingyan; Vora, Mehul; Liu, Jinwei; Huang, Yifan; Mustaine, Elizabeth; and Lai, Vincent S., "Simulating Temporal Dynamics in Cryptocurrency Software Social Networks" (2019). AMCIS 2019 Proceedings. 37.
https://aisel.aisnet.org/amcis2019/data_science_analytics_for_decision_support/data_science_analytics_for_decision_support/37
Simulating Temporal Dynamics in Cryptocurrency Software Social Networks
Cryptocurrency, which relies on decentralized computer software programs to manage monetary transactions, has been associated strongly with illegal money laundering on the dark web. Understanding the social networks of the cryptocurrency software development communities can provide useful investigative leads and business intelligence. However, research in modeling software social networks (SSNs) of cryptocurrency software development is not widely available. This research developed three temporal network models to predict and to simulate the activities of cryptocurrency SSNs over time. The models capture the activity history of the SSNs and dynamically combine the pricing information of the related cryptocurrencies to enhance the accuracy of prediction and fidelity of simulation. We used the models to predict and simulate the SSNs of 83,536 repositories that are related to the three cryptocurrencies: Bitcoin, Monero, and Ethereum. Experimental results show that the model (SEAM) that captures recency / primacy effects of human cognitive processing outperformed other models in metrics pertaining to influence, popularity, and engagement. The research should enhance understanding of online behavior of the SSN community and identify relationship between SSNs and cryptocurrency market.