With the rapid development and widespread popularity of the Internet, the amount of data in social media and networks is growing exponentially, and sentiment analysis for this huge amount of data is very complex but significant. Fine-grained sentiment analysis has become the choice of researchers when dealing with various sentiment analysis tasks. Different from coarse-grained sentiment analysis, which only focuses on emotional polarity, fine-grained sentiment analysis involves emotional polarity and emotional intensity and recipients, providing more specific information about emotions. This paper aims to provide relevant research methods on fine-grained sentiment analysis and apply them to social network texts to analyze the challenges and solutions. This paper will classify fine-grained sentiment analysis from three methods: rule-based, machine learning and deep learning. This research finds that fine-grained sentiment analysis can not only accurately capture the emotions in the text, but also judge the direction and intensity of emotions, and understand different types of emotions in the text more specifically. This is of great help in dealing with more complex texts, such as social network texts. Combining fine-grained sentiment analysis with various large models can solve many challenges and problems when dealing with social network texts.

Show more

In the field of deep learning, current human action recognition algorithms often treat temporal information, spatial information, and background information equally, which leads to limited recognition accuracy. To address this issue, this paper proposes a human action recognition algorithm based on spatiotemporal information interaction. First, a dual-pathway network is proposed to learn spatial and temporal information at different refresh rates. The network includes a sparse pathway operating at a low frame rate to capture spatial semantic information, and a parallel dense pathway operating at a high frame rate to capture temporal motion information. Second, to extract more discriminative features from videos, a cross-dual attention interaction model is introduced to focus on key regions of video segments and explicitly exchange spatiotemporal information between the two pathways. Experimental results show that the proposed algorithm achieves recognition accuracies of 97.6% on the UCF101 dataset and 78.4% on the HMDB51 dataset, outperforming the novel SlowFast algorithm by 1.8% and 1.4%, respectively. Combined with a nighttime image enhancement algorithm based on MDIFE-Net curve estimation, the method achieved an accuracy of 83.2% on the ARID nighttime dataset—an improvement of 22.9% over the performance before image enhancement. This demonstrates the method’s strong potential for real-world nighttime action recognition applications.

Show more

In a closed ecosystem, the correlation between different species and the specific evolution process is important research subject, and the relevant experimental research and field investigations have attracted attention in recent years. In view of the complexity of chaotic systems, this paper aims to introduce and summarize the mathematical description of the aquarium project from multiple aspects, and try to propose a more concise scheme to describe the movement relationship between entities in the ecosystem. The artificial fish school algorithm was used in this research to simulate the response relationship of two species, and the rules of this aquarium were explored. The results show that this ecosystem still evolves to the same result even if the initial parameters are slightly different. This research simplifies the complex calculations in the process and provides new ideas and supplements for understanding the dynamic balance mechanism of closed ecosystems.

Show more

This paper investigates the issue of load disturbances in DC motor speed regulation systems. In response to the limited disturbance rejection capability of traditional PID controllers, a PID controller with feedforward compensation is designed. The study first analyzes the dynamic impact mechanism of sudden load changes on motor speed, and then achieves precise control of the DC motor using PWM technology. Through comparative analysis of simulation and experimental results, the proposed feedforward-compensated PID control demonstrates significant improvements in dynamic response speed, overshoot, and steady-state accuracy compared to traditional PID control. The findings indicate that the proposed control strategy effectively suppresses load disturbances, enhances system stability and robustness, and offers valuable insights for practical engineering applications.

Show more