Wenjun Zheng

With the proliferation of massive open online courses, teaching evaluation texts have become an important source of feedback, providing valuable affective and cognitive insights into students’ learning experiences. Although automated sentiment analysis techniques have been developed, existing methods still struggle to capture domain-specific affective knowledge and to systematically model the affective and cognitive processes inherent in educational contexts. To address these challenges, we propose an affective cognition-enhanced graph convolutional network (ACEGCN), which simulates a human-like sentiment reasoning process by integrating domain-specific affective cognition knowledge with syntactic information. Specifically, we construct a domain-specific affective cognition knowledge graph by aligning sentiment entities with a general knowledge base. The resulting knowledge is then integrated into the input sentence to form a knowledge-enriched sentence tree, from which semantically enhanced dependency graphs are constructed using a multigranularity syntactic masking mechanism. Finally, we employ a graph convolutional network to capture key sentiment paths and model complex semantic dependencies. Experimental results across three real-world teaching evaluation datasets demonstrate that ACEGCN significantly outperforms baseline models, suggesting its potential to enhance teaching feedback analysis and help improve teaching quality.

Mining sentiment from user-generated social media content is crucial for fields such as business decision-making and public opinion analysis. Aspect-based sentiment analysis (ABSA) aims to fine-grainedly identify the sentiment of different entities or attributes in texts. However, existing methods face two challenges: 1) attention mechanisms are prone to interference from aspect-irrelevant words, leading to attention bias; and 2) irrelevant syntactic dependencies are frequently introduced by graph convolutional networks constructed upon dense dependency trees, causing semantic interference. To address these challenges, this article proposes a novel method that combines graph transformer with aspect-oriented graph convolutional network (GT-AGCN), reducing attention bias and semantic interference while modeling global semantics and local syntax. First, a context encoder is constructed to capture contextualized semantic representations of the input text sequences. Second, to efficiently model global semantic information, a syntactically constrained Transformer module is developed to alleviate the inherent attention bias. Meanwhile, to accurately capture local syntax information, an aspect-oriented graph convolutional network is designed to suppress interference from aspect-irrelevant terms. Finally, a classifier is utilized to predict the sentiment polarities of target aspects by combining global semantic features with local syntactic features. Empirical evaluations across five widely-adopted ABSA benchmarks demonstrate that GT-AGCN outperforms state-of-the-art baselines, validating the effectiveness of model.

Healthy living space is the guarantee of high quality of life, dong Village is one of the few minority living space environment adaptation wisdom of the living carrier. By virtue of its healthy ecosystem and the smart wisdom of residents, it is connected with or adapted to the natural environment to create a healthy and livable living space under low-technology conditions. Taking The Gaoxiu Dong Village as the research object, this paper analyzes the environmental adaptation wisdom of the spatial elements such as road system, drum tower space and residential building in Gaoxiu Dong Village, so as to show the spatial form beauty, functional practical beauty and national characteristic beauty of Dong Village. From the perspective of health, it analyzes the adaptive form and mode of human living space, and inherits the wisdom of building with respect to nature and environment.