Complex network of eye movements during rapid automatized naming

Area of Science:

• Neuroscience
• Cognitive Science
• Network Science

Background:

• Visualizing eye-tracking data as time-series can improve understanding of gaze behavior.
• This approach has not been extensively studied in the context of rapid automated naming (RAN).

Purpose of the Study:

• To analyze gaze behavior during RAN using a network-domain perspective.
• To construct a gaze-time-series-based complex network (GCN) from eye-tracking data.
• To extract gaze behavior features by computing GCN's topological parameters without predefined regions of interest.

Main Methods:

• A sample of 98 children (aged 11.50 ± 0.28 years) participated.
• Gaze time-series data were used to construct a GCN for each RAN task.
• Nine topological parameters (average degree, network diameter, characteristic path length, clustering coefficient, global efficiency, assortativity coefficient, modularity, community number, and small-worldness) were computed.

Main Results:

• GCN exhibited assortativity, small-world, and community architecture across RAN tasks.
• Specific topological parameters differentiated between naming numbers/Chinese characters and objects/colors.
• Non-alphanumeric RAN showed distinct GCN properties (higher average degree, global efficiency, small-worldness; lower network diameter, path length, clustering, modularity) compared to alphanumeric RAN.
• Topological parameters were largely independent of traditional eye-movement metrics.

Conclusions:

• The study reveals the architecture and topological parameters of GCN in RAN tasks.
• Task types significantly influence GCN properties.
• This network-based approach provides novel insights into understanding RAN gaze behavior.