<p>Video quality assessment (VQA) plays a critical role in enhancing the user experience on video platforms and services. However, existing deep learning models primarily focus on extracting distorted information, overlooking the relevant insights from the human cognitive system (HCS). DC-HCVQA model, which includes dynamic and static visual feature rectifiers, visual language rectifiers, and memory rectifiers, is designed to tackle VQA complexity from the HCS perspective. The static rectifier innovative leverages the dark channel map to build prior knowledge and enhance static feature extraction; the dynamic rectifier reduces information redundancy and computational complexity by processing video blocks; the visual language rectifier uses a unique scaling factor to merges linguistic and visual data to align with human subjective perception; and the memory rectifier retrieves keyframes as stored memories using memory retrieval and visual features as cues. Our model simulates the modules in HCS more completely and combines them in a unique way compared to other HCS models. Extensive experimental results demonstrate that the DC-HCVQA model achieves state-of-the-art performance across all five popular UGC-VQA datasets, further confirming its effectiveness and robustness.</p>