Scholarlink

<p>The field of sociology requires a deep understanding of complex, multifaceted social data, which often involves ambiguous relationships between entities and concepts. Traditional methods for theory construction may fall short in capturing these intricate dynamics, leading to a demand for advanced analytical tools. To address this, a Recommendation System for Sociology Theory Construction Using Hyperbolic Random Snow Geese Graph coupled Attention Network (Hyp-RS2G-CAN) is proposed. This model leverages hyperbolic graph structures, which are ideal for representing hierarchical and nonlinear relationships in sociological data. It incorporates a Metaphor Processing Method for pre-processing data to interpret abstract expressions better, followed by feature extraction using the Continuous Short-Time Fourier Wavelet Transform (CSTFWF). The Hyp-RS2G-CAN subsequently uses a hyperbolic graph and attention mechanisms to predict the relationships that exist within the data, thus enabling a better premise of the underlying constitutive social provision. Collaborative user-based filtering is then enhanced in the recommendation process to achieve a better result in refining the predicted similarity scores along with making sure the system is making the correct recommendations, which is useful in developing sociological theory. Such an approach is better since it was determined to be more precise (99) with low RMSE (0.6836) hence giving the DSHS-ConAtNet model to be equally more outstanding at building stronger sociological hypotheses.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Recommendation system for sociology theory construction using hyperbolic random snow geese graph coupled attention network

  • 期刊论文

  • 发表日期：

    2025年11月28日

查看全文
  • K. Suresh Kumar,
  • R. Kavitha,
  • Hassan M. Al-Jawahry,
  • Elangovan Muniyandy

摘要

The field of sociology requires a deep understanding of complex, multifaceted social data, which often involves ambiguous relationships between entities and concepts. Traditional methods for theory construction may fall short in capturing these intricate dynamics, leading to a demand for advanced analytical tools. To address this, a Recommendation System for Sociology Theory Construction Using Hyperbolic Random Snow Geese Graph coupled Attention Network (Hyp-RS2G-CAN) is proposed. This model leverages hyperbolic graph structures, which are ideal for representing hierarchical and nonlinear relationships in sociological data. It incorporates a Metaphor Processing Method for pre-processing data to interpret abstract expressions better, followed by feature extraction using the Continuous Short-Time Fourier Wavelet Transform (CSTFWF). The Hyp-RS2G-CAN subsequently uses a hyperbolic graph and attention mechanisms to predict the relationships that exist within the data, thus enabling a better premise of the underlying constitutive social provision. Collaborative user-based filtering is then enhanced in the recommendation process to achieve a better result in refining the predicted similarity scores along with making sure the system is making the correct recommendations, which is useful in developing sociological theory. Such an approach is better since it was determined to be more precise (99) with low RMSE (0.6836) hence giving the DSHS-ConAtNet model to be equally more outstanding at building stronger sociological hypotheses.