The significance of emotion in college music education is becoming increasingly acknowledged. This paper delves into the nexus between music teaching and emotional engagement, advocating for a stronger emphasis on emotional connection within the curriculum. By selectively extracting middle and high-level music features, we aim to intensify the emotional resonance of music pieces. Utilizing the KGIFE model, we gather and analyze students’ emotional responses to music, enabling targeted feedback. Our findings suggest four innovative pathways for evolving music education in higher learning institutions. Comparative analysis reveals that music majors exhibit markedly higher levels of emotion recognition and musical emotion experience than their non-music counterparts, with noticeable differences of 0.1512 and 0.2654 respectively. Furthermore, the study highlights significant improvements in music teaching effectiveness, particularly in run-on training and dialect processing, evidenced by average scores of 3.565 and 3.346. These outcomes underscore the potential of information technology to enhance the delivery and impact of music education in colleges and universities.