Foundations of Temporal Text Networks

Area of Science:

• Network Science
• Computational Social Science
• Information Science

Background:

• Human information networks involve actors, exchanged information (text content), and temporal dynamics.
• Existing research often studies these elements in isolation or pairs, leading to fragmented understanding.
• Current studies of all three elements frequently employ non-transferable, ad hoc models.

Purpose of the Study:

• To present a unified, simple, expressive, and extensible model for temporal text networks.
• To provide a common framework for analyzing diverse network types and tasks.
• To demonstrate how this model facilitates the application of established analysis methods.

Main Methods:

• Development of a novel, unified model for temporal text networks.
• Proposing simple procedures to generate network views from the model.
• Leveraging existing data mining and multilayer network mining techniques.

Main Results:

• The proposed model effectively integrates actors, text content, and time.
• Generated network views allow direct application of established analytical methods.
• The model's flexibility supports various network analysis tasks and domains.

Conclusions:

• The presented model offers a foundational approach for understanding complex temporal text networks.
• This unified framework promotes cross-contextual analysis and method transfer.
• The model simplifies the application of advanced data mining techniques to temporal information networks.