A vision sensing-enhanced knowledge graph inference method for a healthy operation index in higher education

Area of Science:

• Educational Technology
• Artificial Intelligence
• Data Science

Background:

• Assessing the healthy operation of higher education institutions is crucial for strategic development.
• Existing data-driven methods may lack the depth to capture complex operational nuances.
• Knowledge mapping offers a powerful visualization and analytical tool for complex systems.

Purpose of the Study:

• To propose a knowledge mapping-based construction method for inferring the healthy operation index in higher education (HOI-HE).
• To develop a vision sensing-enhanced knowledge graph method for HOI-HE evaluation.
• To create a digital evaluation platform for HOI-HE value through integrated functional modules.

Main Methods:

• An improved named entity identification and relationship extraction method using BERT vision sensing pre-training.
• A multi-decision model-based knowledge graph for inferring HOI-HE scores via multi-classifier ensemble learning.
• Integration of knowledge extraction, relational reasoning, and triadic quality evaluation modules.

Main Results:

• The vision sensing-enhanced knowledge graph method effectively infers HOI-HE scores.
• The proposed method demonstrates benefits over purely data-driven approaches.
• Experimental results in simulated scenes show successful HOI-HE evaluation and latent risk discovery.

Conclusions:

• The developed vision sensing-enhanced knowledge inference method provides a robust digital evaluation platform for HOI-HE.
• This approach enhances the accuracy and depth of HOI-HE assessment compared to traditional methods.
• The method has the potential to identify subtle risks within higher education operations.