At present, assessing the impact of university English classes is a key component of evaluating the quality of higher education instruction. Based on the evaluation results, instructors can better understand the impact of their lessons and make necessary adjustments to their current pedagogical practices and curriculum. Administrators at the school may use the test scores to gauge English instructors’ all-around competence in the classroom. This paper presents an evaluation model that improves the evaluation effect, uses the global search method to realize individual information sharing and transmission, updates the speed and position of the next iteration, and allows particles to search for individual extreme values and global extrema in the population. The evaluation system is built using particle swarm algorithm and support vector machine. The evaluation system is used to assess the effectiveness of English teaching in universit By using this approach, the impact of English language instruction in higher education may be scientifically evaluated, with less influence from subjective opinions and chance. The findings indicate that the algorithm is capable of achieving a reasonable assessment effect while also drastically improving evaluation accuracy and decreasing evaluation cycle time.