Simulating emergence of novelties using agent-based models.

Area of Science:

• Computational Social Science
• Network Science
• Agent-Based Modeling

Background:

• Social networks underpin online services like social networking sites (SNS) and games.
• Existing models fail to capture real-world network behaviors such as "waves of novelty."

Purpose of the Study:

• To develop a novel agent-based model that incorporates context information for more accurate social network evolution.
• To better understand and reproduce the structure and growth dynamics of complex networks.

Main Methods:

• Introduced context information via labels based on agent appearance timing and relationships.
• Modified agent selection probabilities to account for novelty and preferential attachment.
• Validated the model against real-world data using ten distinct metrics.

Main Results:

• The new model successfully replicates observed social network behaviors, including "waves of novelty."
• Contextual labels and modified selection probabilities enhance model accuracy.
• The model demonstrates closer resemblance to real data compared to previous approaches.

Conclusions:

• The proposed model offers a more accurate representation of social network growth and evolution.
• Understanding these dynamics is crucial for effective user acquisition and retention strategies.
• This research advances the field of computational social science and network analysis.