Application of Emotional Interaction Analysis Based on Deep Learning in Psychological Defense of Network Security in College Innovation and Entrepreneurship Education

Abstract

This research explores the use of emotional interaction analysis, supplemented with deep learning methods, under the umbrella of psychological defense mechanisms for bolstering network security in college entrepreneurial and innovation teaching. The research creates a complete psychological defense approach combining emotional interaction analysis with deep learning techniques for reinforcing psychological resilience and improving learners' understanding of network security. The system implemented utilises a hybrid approach combining Convolutional Neural Networks (CNN) with Long Short-Term Memory (LSTM) networks, utilizing multi-layered artificial neural frameworks for emotional interaction data-related feature extractions. The approach supports the measurement of emotional response and behavioral trends, hence enabling emotional state and psychological vulnerability identification. The experimental results corroborate the efficacy of the system, with accuracy levels of 94.7% for emotional state identification and 89.7% for psychological vulnerability identification, hence surpassing traditional approaches. Practically, the system enhances learners' psychological defense mechanisms and perception of security, with psychological resilience improvement of 36.6% and accuracy of security response of 94.3%. The study brings forth a novel fusion of emotional interaction and deep learning, hence giving rise to a psychological defense model applicable in network security teaching in entrepreneurial and innovative higher education environments in colleges. The research also contributes considerably towards pedagogic knowledge and theoretical frameworks, hence giving rise to a novel teaching approach for network security incorporating psychological defense mechanisms.