With the rapid development of information technology, the internet has emerged as a pivotal driving force in reshaping higher education paradigms. This paper delves into clustering algorithms and proposes an enhanced version, exploring how this enhanced clustering algorithm can be applied to blended teaching of digital electronic technology courses within an "internet+" framework. The objective is to refine teaching methodologies and elevate students' learning capabilities. The proposed approach initially mines student learning behavior patterns to comprehensively understand their learning states and accurately assess their behaviors, thereby optimizing teaching practices. Subsequently, evaluations and recommendations for enhancing the teaching process are formulated. Experimental results underscore the ability to classify students into distinct categories according to their learning behavior habits, guiding educators to tailor teaching plans more effectively, enhance teaching quality, and ultimately benefit the sustained development of blended learning approaches.