Sentiment analysis, also known as opinion mining, is a crucial branch of Natural language processing, which focuses on recognizing, extracting, and quantifying sentiment tendencies, emotional intensity and specific emotion types in textual data. With the rapid development of the internet and communication, analyzing sentiment contained in textual data becomes important and crucial for understanding public opinion, consumer behavior, and emotional trends. This paper provides a comprehensive review of sentiment analysis in the range of its application, evolution, task types, methodology and future development by analyzing the literature of this field. Sentiment analysis has developed from traditional lexicon-based methods to modern deep learning methods like CNN, RNN and transformer model, which have significantly improved accuracy and robustness. This paper also discussed challenges in sentiment analysis like sarcasm detection and cross-lingual analysis, and proposed potential solutions. The findings aim to provide comprehensive insight for researchers and contribute to innovations in sentiment analysis.

Show more

Since the inception and evolution of artificial intelligence, control systems, particularly automated pathfinding, have constituted a pivotal area of significance. It is reflected in today's world of life in all aspects of autonomous driving, rescue robots and so on. Algorithms as the basis for the realization of this field exist in thousands of situations, which are reflected in different logic, different computational efficiency, different time and space complexity, etc. We will introduce the development history and basic principles of Dijkstra’s algorithm and Rapidly-exploring Random Trees algorithm, and analyze the advantages and disadvantages of each of these two algorithms to determine the domains in which they are applicable. In this paper, we will use MATLAB to set up a test environment and will investigate the effect of different environments comparing Dijkstra’s algorithm and RRT algorithm on the automatic control system in artificial intelligence. In addition, some experimental data and icons will be cited to support the experimental results by comparing the algorithms in terms of distance and efficiency. It is concluded that Dijkstra’s algorithm will be more suitable for static and low dimensional environments, while Rapidly-exploring Random Trees algorithm will be suitable for more complex and high dimensional environments.

Show more

This study develops a novel bi-level optimization framework incorporating carbon pricing mechanisms for integrated electricity-gas systems, enabling cost-effective low-carbon energy dispatch. The proposed approach employs a carbon emission flow model to quantify nodal carbon potentials and track emission pathways across energy networks. The optimization architecture consists of two interconnected layers: the upper layer minimizes power grid and natural gas network operational costs, while the lower layer optimizes energy procurement, equipment maintenance, and carbon trading expenditures. The alternating direction method of multipliers (ADMM) algorithm is implemented to solve this complex optimization problem. Case study results demonstrate the framework's effectiveness in balancing economic and environmental objectives: system-wide carbon emissions were reduced by 4.77 tCO₂, with power procurement costs decreasing by 6,533.45 yuan. Although natural gas expenses increased by 6,569.07 yuan due to the carbon trading mechanism, the overall framework achieved an improved equilibrium between operational efficiency and sustainable energy practices, providing valuable insights for low-carbon energy system optimization.

Show more

The ethical dilemma of cybersecurity in the era of big data presents a multi-dimensional outbreak trend. As the speed of technological change far exceeds the update of laws and regulations, the crisis of technological trust and social ethical conflicts are intertwined, forcing the reconstruction of the global governance system. This study comprehensively deconstructs the ethical disputes in the current field of cybersecurity and summarizes five core contradictions, including the zero-sum game between monitoring rights and privacy rights, the systematic spread of algorithmic discrimination, the new hegemony of data colonialism, the gap between rights and responsibilities of vulnerability disclosure, and the failure of the AI application attribution mechanism. Based on the comprehensive analysis and systematic integration of multi-dimensional fragmented cases, this study proposes a hierarchical and progressive governance paradigm: the technical governance layer pragmatically improves the existing system, the institutional coordination layer links multiple mechanisms, and the cultural identity layer improves the digital citizen literacy. By balancing the dual logic of technological innovation and value constraints, this framework provides an operational governance path for cybersecurity regulatory departments to optimize ethical risk assessment tools, Internet companies to establish algorithm audit committees, and technology research and development institutions to improve ethical embedded design. It has important theoretical reference value for building a humanistic-oriented digital civilization order.

Show more