The extensive application of algorithm recommendation technology not only meets the information needs of network users but also leads to the emergence of “information cocoons”. On the basis of summarizing three generating mechanisms, namely, the theory of technological innovation, the theory of interest-driven, and the theory of emotional identity, this paper constructs a game model of bilateral evolution between information platforms and network users and simulates the influence path of key factors on the evolution of both parties’ main strategies. The research shows that algorithm recommendation technology is the root of “information cocoons” in the algorithm era. As the algorithm technology matures day by day, the cost of using algorithm recommendations on information platforms and the loss cost of accepting algorithm recommendations by network users are constantly decreasing, which causes the information platforms and network users’ strategy choice for algorithm recommendation to evolve from {give up and conflict} to {use and accept}, and finally leads to the long-term existence of “information cocoons”.

Purpose: This paper aims to clarify the impact of information asymmetry on users' payment rates and examine the role of perceived uncertainty (PU) and acceptable price (AP) in the relationship between information asymmetry and users' payment rates. Design/methodology/approach: To test the influences of information asymmetry on users' payment rates, this paper collects 18,489 transaction data from the Chinese knowledge payment platform Zhihu with a Python crawler. This paper constructs a mediation model to define the relationship between information asymmetry and users' payment rates by introducing PU and AP as the mediators. Findings: Information asymmetry negatively affects users' payment rates. In addition, PU and AP mediate the information asymmetry in users' payment rates bond. Research limitations/implications: This study only explores the mediators of the information asymmetry users’ payment rates bond, ignoring the effect of potential moderators, which would be an important direction for future research. Practical implications: The findings of this paper suggest that information communication is essential in knowledge market transactions. Knowledge providers, as well as knowledge platforms, should enhance information exchange with consumers in order to increase product sales. Social implications: This paper provides a new perspective for understanding how information asymmetry affects users' payment rates and helps to guide suppliers to improve product quality. The research framework of this paper is universal to a certain extent. Originality/value: This paper is one of the first to propose using PU and AP to construct a mediation model to study the information asymmetry between users' payment rates relationship. It provides a new perspective for understanding the channel of information asymmetry in customer behavior.

Purpose: This paper aims to propose a solution for dissolving the “privacy paradox” in social networks, and explore the feasibility of adopting a synergistic mechanism of “deep-learning algorithms” and “differential privacy algorithms” to dissolve this issue. Design/methodology/approach: To validate our viewpoint, this study constructs a game model with two algorithms as the core strategies. Findings: The “deep-learning algorithms” offer a “profit guarantee” to both network users and operators. On the other hand, the “differential privacy algorithms” provide a “security guarantee” to both network users and operators. By combining these two approaches, the synergistic mechanism achieves a balance between “privacy security” and “data value”. Practical implications: The findings of this paper suggest that algorithm practitioners should accelerate the innovation of algorithmic mechanisms, network operators should take responsibility for users’ privacy protection, and users should develop a correct understanding of privacy. This will provide a feasible approach to achieve the balance between “privacy security” and “data value”. Originality/value: These findings offer some insights into users’ privacy protection and personal data sharing.