Applying Answer Set Programming for Knowledge-Based Link Prediction on Social Interaction Networks.

Area of Science:

• Social Network Analysis
• Artificial Intelligence
• Data Mining

Background:

• Link prediction in social networks aims to forecast future connections based on current data.
• Predicting future links is challenging due to limited historical information.
• Integrating external domain knowledge can potentially improve prediction accuracy.

Purpose of the Study:

• To investigate the use of additional domain knowledge for improving social network link prediction.
• To formalize domain knowledge for social network analysis using Answer Set Programming (ASP).
• To explore explanation-aware prediction approaches enabled by ASP.

Main Methods:

• Utilized Answer Set Programming (ASP) to formalize domain knowledge for graph analysis.
• Applied ASP for link prediction based on node proximity.
• Enhanced link prediction by incorporating background knowledge, such as shared educational backgrounds.

Main Results:

• Demonstrated that incorporating domain knowledge, like common educational backgrounds, can enhance link prediction accuracy.
• Showcased the capability of ASP to integrate background information into prediction models.
• Validated the intuition that shared features imply common interests for link prediction.

Conclusions:

• Answer Set Programming (ASP) provides a robust framework for integrating domain knowledge into social network link prediction.
• The proposed ASP-based approach enhances prediction by leveraging background information and enabling explanation-aware predictions.
• This method offers a promising direction for more accurate and interpretable social network analysis.