Pub Date : 2025-03-20DOI: 10.1016/j.sasc.2025.200221
Weina Yu
In order to address the limitations of traditional models in generating music styles, a multi style music generation model has been designed to support music teaching. The main contribution of the research is the introduction of a Multi style chord music generation network to enhance the adaptability and innovative generation ability of the model to different music styles. The weight of different music styles is adjusted through a style transfer mechanism to achieve seamless transition of chord styles. The experimental results show that the loss value of the research method is 0.16, and the accuracy of the model's note recognition is 81.68%, both of which reach a high level. The accuracy, recall, and F1 score of the research method for music sequence recognition are 95.16%, 92.53%, and 0.948, respectively, all of which are better than the comparative models. This indicates that the research method has better flexibility in music generation and stronger ability to generate multi style music. Research can aid with the generation of multi style music in music teaching.
为了解决传统模型在生成音乐风格方面的局限性，我们设计了一个多风格音乐生成模型来支持音乐教学。该研究的主要贡献在于引入了多风格和弦音乐生成网络，以增强模型对不同音乐风格的适应性和创新生成能力。通过风格转换机制调整不同音乐风格的权重，实现和弦风格的无缝转换。实验结果表明，研究方法的损失值为 0.16，模型的音符识别准确率为 81.68%，均达到较高水平。研究方法对音乐序列识别的准确率、召回率和 F1 分数分别为 95.16%、92.53% 和 0.948，均优于对比模型。这表明该研究方法在音乐生成方面具有更好的灵活性和更强的生成多风格音乐的能力。该研究有助于音乐教学中多风格音乐的生成。
{"title":"The construction of improved GCA multi-style music generation model for music intelligent teaching classroom","authors":"Weina Yu","doi":"10.1016/j.sasc.2025.200221","DOIUrl":"10.1016/j.sasc.2025.200221","url":null,"abstract":"<div><div>In order to address the limitations of traditional models in generating music styles, a multi style music generation model has been designed to support music teaching. The main contribution of the research is the introduction of a Multi style chord music generation network to enhance the adaptability and innovative generation ability of the model to different music styles. The weight of different music styles is adjusted through a style transfer mechanism to achieve seamless transition of chord styles. The experimental results show that the loss value of the research method is 0.16, and the accuracy of the model's note recognition is 81.68%, both of which reach a high level. The accuracy, recall, and F1 score of the research method for music sequence recognition are 95.16%, 92.53%, and 0.948, respectively, all of which are better than the comparative models. This indicates that the research method has better flexibility in music generation and stronger ability to generate multi style music. Research can aid with the generation of multi style music in music teaching.</div></div>","PeriodicalId":101205,"journal":{"name":"Systems and Soft Computing","volume":"7 ","pages":"Article 200221"},"PeriodicalIF":0.0,"publicationDate":"2025-03-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143768968","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-19DOI: 10.1016/j.sasc.2025.200223
Lei Zou
The intelligent Internet of Things has played a certain role in the tourism industry. In the evaluation of tourist destination competitiveness, the text evaluation and images of tourist destination can be collected through the Internet of Things. Among them, text data processing is relatively simple, but image and video processing is more difficult, and different data sources will lead to problems such as the decline of federated learning algorithms. In order to improve the data processing problem in the evaluation of tourist destination competitiveness and to solve the problem of unbalanced utilization of computing and communication resources caused by system heterogeneity, this paper proposes an adaptive asynchronous aggregation method Adaptive asynchronous aggregation method based on outdated threshold control (HiFedCNM) based on obsolescence threshold control. The experimental results show that the algorithm outperforms some existing excellent algorithms in model training accuracy, computational efficiency, communication efficiency and system cost. In addition, this paper proposes a conceptual model of tourism destination competitiveness. Through the case study, it can be seen that the model proposed in this paper can play a certain role in the analysis of tourist destination competitiveness. At the same time, the model method proposed in this paper can provide a reliable reference for the subsequent heterogeneous fusion of tourism Internet of Things data, and can provide a reliable method for evaluating the competitiveness of tourism destinations.
{"title":"Tourism destination competitiveness evaluation model integrating multi-source big data and machine learning","authors":"Lei Zou","doi":"10.1016/j.sasc.2025.200223","DOIUrl":"10.1016/j.sasc.2025.200223","url":null,"abstract":"<div><div>The intelligent Internet of Things has played a certain role in the tourism industry. In the evaluation of tourist destination competitiveness, the text evaluation and images of tourist destination can be collected through the Internet of Things. Among them, text data processing is relatively simple, but image and video processing is more difficult, and different data sources will lead to problems such as the decline of federated learning algorithms. In order to improve the data processing problem in the evaluation of tourist destination competitiveness and to solve the problem of unbalanced utilization of computing and communication resources caused by system heterogeneity, this paper proposes an adaptive asynchronous aggregation method Adaptive asynchronous aggregation method based on outdated threshold control (HiFedCNM) based on obsolescence threshold control. The experimental results show that the algorithm outperforms some existing excellent algorithms in model training accuracy, computational efficiency, communication efficiency and system cost. In addition, this paper proposes a conceptual model of tourism destination competitiveness. Through the case study, it can be seen that the model proposed in this paper can play a certain role in the analysis of tourist destination competitiveness. At the same time, the model method proposed in this paper can provide a reliable reference for the subsequent heterogeneous fusion of tourism Internet of Things data, and can provide a reliable method for evaluating the competitiveness of tourism destinations.</div></div>","PeriodicalId":101205,"journal":{"name":"Systems and Soft Computing","volume":"7 ","pages":"Article 200223"},"PeriodicalIF":0.0,"publicationDate":"2025-03-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143704195","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-17DOI: 10.1016/j.sasc.2025.200220
Liang Zhou, Mingyun Tang, Jian Liu
The quality of MOOCs is influenced by multiple factors. This paper proposes a new method, I-DEMATEL-ISM, which combines intuitionistic fuzzy sets with DEMATEL (Decision testing and evaluation laboratory) and ISM (Interpretive structural modeling), to analyze the relationships between these factors and identify strategies for improving MOOC quality. In the first, six key factors were selected through literature research and expert consultations. Then, the interrelationships between the six factors were analyzed. Finally, interpretive structural modeling was established. The analysis revealed that course teachers are fundamental factors, while course content and management are intermediate factors. Learning platforms, tasks, and materials are surface factors. Improving surface factors can enhance MOOC quality in the short term, while improving intermediate and fundamental factors can create a sustainable cycle of quality improvement.
mooc的质量受到多种因素的影响。本文提出了一种新的方法I-DEMATEL-ISM，该方法将直觉模糊集与DEMATEL （Decision testing and evaluation laboratory）和ISM （Interpretive structural modeling）相结合，分析这些因素之间的关系，确定提高MOOC质量的策略。首先，通过文献研究和专家咨询，选择了六个关键因素。然后，分析了六个因素之间的相互关系。最后，建立了解释结构模型。分析表明，课程教师是基础因素，课程内容和管理是中间因素。学习平台、任务和材料是表面因素。改善表面因素可以在短期内提高MOOC质量，而改善中间和基础因素可以形成一个持续的质量改善循环。
{"title":"Analysis of factors influencing MOOC quality based on I-DEMATEL-ISM method","authors":"Liang Zhou, Mingyun Tang, Jian Liu","doi":"10.1016/j.sasc.2025.200220","DOIUrl":"10.1016/j.sasc.2025.200220","url":null,"abstract":"<div><div>The quality of MOOCs is influenced by multiple factors. This paper proposes a new method, I-DEMATEL-ISM, which combines intuitionistic fuzzy sets with DEMATEL (Decision testing and evaluation laboratory) and ISM (Interpretive structural modeling), to analyze the relationships between these factors and identify strategies for improving MOOC quality. In the first, six key factors were selected through literature research and expert consultations. Then, the interrelationships between the six factors were analyzed. Finally, interpretive structural modeling was established. The analysis revealed that course teachers are fundamental factors, while course content and management are intermediate factors. Learning platforms, tasks, and materials are surface factors. Improving surface factors can enhance MOOC quality in the short term, while improving intermediate and fundamental factors can create a sustainable cycle of quality improvement.</div></div>","PeriodicalId":101205,"journal":{"name":"Systems and Soft Computing","volume":"7 ","pages":"Article 200220"},"PeriodicalIF":0.0,"publicationDate":"2025-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143685574","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-17DOI: 10.1016/j.sasc.2025.200218
Yajing Sun
With the rapid development of big data and artificial intelligence technology, personalized learning has attracted significant attention in education. This study focuses on constructing and refining personalized learning paths for English learners by integrating the sparrow search algorithm (SSA) with the long short-term memory (LSTM) model. SSA, an intelligent optimization algorithm, exhibits robust global search capabilities and swift convergence, while the LSTM model excels in processing time series data. This study employs the LSTM model to analyze English learners' behavior data, subsequently optimizing the LSTM model's hyperparameters using SSA to enhance prediction accuracy and generalization. Results demonstrate that the personalized learning path generated by the SSA-LSTM model outperforms the traditional LSTM model and other comparative models across multiple evaluation metrics, offering a more precise prediction of learners' needs and providing educators with a scientific and efficient personalized teaching tool.
{"title":"Construction and optimization of personalized learning paths for English learners based on SSA-LSTM model","authors":"Yajing Sun","doi":"10.1016/j.sasc.2025.200218","DOIUrl":"10.1016/j.sasc.2025.200218","url":null,"abstract":"<div><div>With the rapid development of big data and artificial intelligence technology, personalized learning has attracted significant attention in education. This study focuses on constructing and refining personalized learning paths for English learners by integrating the sparrow search algorithm (SSA) with the long short-term memory (LSTM) model. SSA, an intelligent optimization algorithm, exhibits robust global search capabilities and swift convergence, while the LSTM model excels in processing time series data. This study employs the LSTM model to analyze English learners' behavior data, subsequently optimizing the LSTM model's hyperparameters using SSA to enhance prediction accuracy and generalization. Results demonstrate that the personalized learning path generated by the SSA-LSTM model outperforms the traditional LSTM model and other comparative models across multiple evaluation metrics, offering a more precise prediction of learners' needs and providing educators with a scientific and efficient personalized teaching tool.</div></div>","PeriodicalId":101205,"journal":{"name":"Systems and Soft Computing","volume":"7 ","pages":"Article 200218"},"PeriodicalIF":0.0,"publicationDate":"2025-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143814899","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-17DOI: 10.1016/j.sasc.2025.200219
Xiyuan Gao , Ruohan Gao
With the growth of diverse music information processing needs, music signal recognition technology has become more and more important in music education and music industry. In this study, a music signal recognition aider using convolutional neural network is proposed, and firstly, the logarithmic frequency domain filter bank and double-layer ReLU network are used to extract the pitch features in the music signal. Subsequently, the benchmark convolutional neural network model is constructed, and the constant Q transform is used to process the obtained features to generate a harmonic sequence matrix. Finally, a two-level classification model strategy is used to improve instrument signal recognition. In terms of pitch feature extraction, the accuracy of the logarithmic frequency domain filter group was 74.59 % and 77.03 % respectively under the frame length of 2048 and 8192, which was more effective than the double-layer ReLU network. Experimental results based on different harmonic mapping matrix levels showed that these harmonic mapping matrices had a significant impact on the recall and accuracy of different musical instruments, such as the F1 score of 0.936 for pianos. In the verification of the two-level classification model, the overall accuracy was improved from 0.848 to 0.880 of the benchmark model, which proved the effective improvement of multi-instrument music signal generalization recognition. The research contribution is to improve the ability of pitch feature extraction and establish a more efficient classification model for multi-instrument music signals. These contributions fill the research gap in extracting the pitch and part information of multiple instruments quickly and accurately in complex music works, provide powerful technical support for music analysis and understanding in music education, and innovatively promote the development of music information retrieval technology.
{"title":"Music signal recognition aids based on convolutional neural networks in music education","authors":"Xiyuan Gao , Ruohan Gao","doi":"10.1016/j.sasc.2025.200219","DOIUrl":"10.1016/j.sasc.2025.200219","url":null,"abstract":"<div><div>With the growth of diverse music information processing needs, music signal recognition technology has become more and more important in music education and music industry. In this study, a music signal recognition aider using convolutional neural network is proposed, and firstly, the logarithmic frequency domain filter bank and double-layer ReLU network are used to extract the pitch features in the music signal. Subsequently, the benchmark convolutional neural network model is constructed, and the constant Q transform is used to process the obtained features to generate a harmonic sequence matrix. Finally, a two-level classification model strategy is used to improve instrument signal recognition. In terms of pitch feature extraction, the accuracy of the logarithmic frequency domain filter group was 74.59 % and 77.03 % respectively under the frame length of 2048 and 8192, which was more effective than the double-layer ReLU network. Experimental results based on different harmonic mapping matrix levels showed that these harmonic mapping matrices had a significant impact on the recall and accuracy of different musical instruments, such as the F1 score of 0.936 for pianos. In the verification of the two-level classification model, the overall accuracy was improved from 0.848 to 0.880 of the benchmark model, which proved the effective improvement of multi-instrument music signal generalization recognition. The research contribution is to improve the ability of pitch feature extraction and establish a more efficient classification model for multi-instrument music signals. These contributions fill the research gap in extracting the pitch and part information of multiple instruments quickly and accurately in complex music works, provide powerful technical support for music analysis and understanding in music education, and innovatively promote the development of music information retrieval technology.</div></div>","PeriodicalId":101205,"journal":{"name":"Systems and Soft Computing","volume":"7 ","pages":"Article 200219"},"PeriodicalIF":0.0,"publicationDate":"2025-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143697419","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-13DOI: 10.1016/j.sasc.2025.200207
Chao Jiang , Ting Jiang , Ziyi Lai , Xinwei Ou , Libin Gou
Background
China encounters difficulties in balancing economic development and cultural protection in the tourism sector, especially as urbanization and commercialization influence heritage integrity. This research concentrates on improving wellness-oriented travel itineraries in China's cultural and recreational tourism field.
Objectives
The main objective is to create a big data-driven framework that offers tangible, optimized itineraries to improve tourist fulfillment and assist sustainable tourism development.
Methods
The analysis of tourist data, customer desires, and travel habits employs a mixture of collaborative filtering methods, topic models, and vector space models. The research describes the development of cultural and innovative tourism, and it uses big data analytics to create improved wellness-focused travel paths.
Results
The suggested big data-driven framework enhances travel path optimization by tailoring itineraries to tourist patterns and desires. Comparative performance evaluation shows that the novel technique outperforms previous methods with a recall of 98 %, an F1 score of 98.5 %, an accuracy of 98 %, and a precision of 97 %. These findings support the technique's efficacy in improving wellness tourism services.
Conclusion
This research tackles major obstacles in China's cultural tourism industry by implementing a tangible big data-driven itinerary optimization framework. The findings demonstrate that combining wellness-focused itineraries with cultural and creative components improves tourist fulfillment and guarantees long-term development. This method offers tourism planners useful knowledge for balancing cultural preservation and economic advancement.
{"title":"Optimization of the wellness path in China's cultural and recreational tourism industry: A data-driven framework for health-focused travel plans","authors":"Chao Jiang , Ting Jiang , Ziyi Lai , Xinwei Ou , Libin Gou","doi":"10.1016/j.sasc.2025.200207","DOIUrl":"10.1016/j.sasc.2025.200207","url":null,"abstract":"<div><h3>Background</h3><div>China encounters difficulties in balancing economic development and cultural protection in the tourism sector, especially as urbanization and commercialization influence heritage integrity. This research concentrates on improving wellness-oriented travel itineraries in China's cultural and recreational tourism field.</div></div><div><h3>Objectives</h3><div>The main objective is to create a big data-driven framework that offers tangible, optimized itineraries to improve tourist fulfillment and assist sustainable tourism development.</div></div><div><h3>Methods</h3><div>The analysis of tourist data, customer desires, and travel habits employs a mixture of collaborative filtering methods, topic models, and vector space models. The research describes the development of cultural and innovative tourism, and it uses big data analytics to create improved wellness-focused travel paths.</div></div><div><h3>Results</h3><div>The suggested big data-driven framework enhances travel path optimization by tailoring itineraries to tourist patterns and desires. Comparative performance evaluation shows that the novel technique outperforms previous methods with a recall of 98 %, an F1 score of 98.5 %, an accuracy of 98 %, and a precision of 97 %. These findings support the technique's efficacy in improving wellness tourism services.</div></div><div><h3>Conclusion</h3><div>This research tackles major obstacles in China's cultural tourism industry by implementing a tangible big data-driven itinerary optimization framework. The findings demonstrate that combining wellness-focused itineraries with cultural and creative components improves tourist fulfillment and guarantees long-term development. This method offers tourism planners useful knowledge for balancing cultural preservation and economic advancement.</div></div>","PeriodicalId":101205,"journal":{"name":"Systems and Soft Computing","volume":"7 ","pages":"Article 200207"},"PeriodicalIF":0.0,"publicationDate":"2025-03-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143685575","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-12DOI: 10.1016/j.sasc.2025.200212
Ying Wang
Personalized federated learning (PFL) enables collaborative model training across devices while adapting to heterogeneous data, but faces resource constraints on edge devices. Combining PFL with pruning techniques helps address these constraints. A challenge is that one-size-fits-all pruning strategies may ignore the varying importance of parameters for local data. To overcome this, we propose PLFL, a novel personalized lightweight federated learning framework. PLFL uses a hypernetwork at the server level to deliver personalized local models to clients and incorporates a federated pruning mechanism tailored to parameter importance, ensuring optimal performance and maintaining personalization. Experimental results show that PLFL achieves higher accuracy with lower computational costs and fewer parameters compared to state-of-the-art methods on heterogeneous datasets.
{"title":"Personalized lightweight federated learning for efficient and private model training in heterogeneous data environments","authors":"Ying Wang","doi":"10.1016/j.sasc.2025.200212","DOIUrl":"10.1016/j.sasc.2025.200212","url":null,"abstract":"<div><div>Personalized federated learning (PFL) enables collaborative model training across devices while adapting to heterogeneous data, but faces resource constraints on edge devices. Combining PFL with pruning techniques helps address these constraints. A challenge is that one-size-fits-all pruning strategies may ignore the varying importance of parameters for local data. To overcome this, we propose PLFL, a novel personalized lightweight federated learning framework. PLFL uses a hypernetwork at the server level to deliver personalized local models to clients and incorporates a federated pruning mechanism tailored to parameter importance, ensuring optimal performance and maintaining personalization. Experimental results show that PLFL achieves higher accuracy with lower computational costs and fewer parameters compared to state-of-the-art methods on heterogeneous datasets.</div></div>","PeriodicalId":101205,"journal":{"name":"Systems and Soft Computing","volume":"7 ","pages":"Article 200212"},"PeriodicalIF":0.0,"publicationDate":"2025-03-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143685571","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The tourism industry is a major economic sector worldwide, significantly contributing to job creation and GDP growth. However, the rapid expansion of this industry, along with rising environmental and social concerns, underscores the critical need for sustainable strategies. This paper presents a novel multi-objective game theory model that simultaneously optimizes profitability and sustainability in the tourism supply chain. The key innovation of this study lies in the integration of game theory with an artificial neural network (ANN) to predict customer demand, effectively capturing nonlinear consumer behaviors and enabling more accurate decision-making. The model analyzes the dynamic interactions between tour operators and local service providers, identifying Nash Equilibrium outcomes where no player can improve profitability through unilateral strategy adjustments. Additionally, the study introduces a comprehensive approach to government subsidies, evaluating their effectiveness in enhancing sustainability incentives and overall profitability. A detailed sensitivity analysis is conducted to examine how variations in pricing, sustainability efforts, and subsidy rates influence profit margins. Another distinctive contribution of this research is its emphasis on human resource management, highlighting how employee training, green organizational culture, and financial incentives can improve productivity and support sustainability initiatives. The results demonstrate that collaborative strategies, such as resource sharing and joint sustainability efforts between tour operators and local providers, significantly increase profitability. The findings further indicate that a combination of optimal pricing, maximum sustainability efforts, and full government subsidies yields the highest total profit of 6,395 units. Overall, this research offers strategic guidelines for pricing, human resource development, and subsidy policies, providing a robust framework for achieving both profitability and sustainability in the tourism supply chain.
{"title":"A multi-objective game theory model for sustainable profitability in the tourism supply chain: Integrating human resource management and artificial neural networks","authors":"Amirhossein Torkabadi , Mobina Mousapour Mamoudan , Babek Erdebilli , Amir Aghsami","doi":"10.1016/j.sasc.2025.200217","DOIUrl":"10.1016/j.sasc.2025.200217","url":null,"abstract":"<div><div>The tourism industry is a major economic sector worldwide, significantly contributing to job creation and GDP growth. However, the rapid expansion of this industry, along with rising environmental and social concerns, underscores the critical need for sustainable strategies. This paper presents a novel multi-objective game theory model that simultaneously optimizes profitability and sustainability in the tourism supply chain. The key innovation of this study lies in the integration of game theory with an artificial neural network (ANN) to predict customer demand, effectively capturing nonlinear consumer behaviors and enabling more accurate decision-making. The model analyzes the dynamic interactions between tour operators and local service providers, identifying Nash Equilibrium outcomes where no player can improve profitability through unilateral strategy adjustments. Additionally, the study introduces a comprehensive approach to government subsidies, evaluating their effectiveness in enhancing sustainability incentives and overall profitability. A detailed sensitivity analysis is conducted to examine how variations in pricing, sustainability efforts, and subsidy rates influence profit margins. Another distinctive contribution of this research is its emphasis on human resource management, highlighting how employee training, green organizational culture, and financial incentives can improve productivity and support sustainability initiatives. The results demonstrate that collaborative strategies, such as resource sharing and joint sustainability efforts between tour operators and local providers, significantly increase profitability. The findings further indicate that a combination of optimal pricing, maximum sustainability efforts, and full government subsidies yields the highest total profit of 6,395 units. Overall, this research offers strategic guidelines for pricing, human resource development, and subsidy policies, providing a robust framework for achieving both profitability and sustainability in the tourism supply chain.</div></div>","PeriodicalId":101205,"journal":{"name":"Systems and Soft Computing","volume":"7 ","pages":"Article 200217"},"PeriodicalIF":0.0,"publicationDate":"2025-03-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143619845","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-08DOI: 10.1016/j.sasc.2025.200216
Daocai Wang
This paper constructs two cross-domain recommendation models based on the perspective of user sharing and non-sharing, both of which rely on intensive convolutional networks and attention mechanisms. This research introduces lightweight Dense Net and fine-grained pruning for model optimization. Lightweight Dense Net retains the core advantages by optimizing the repeat structure while reducing redundant parameters. Compared with the original network, the accuracy loss is not >2 %, the number of parameters is reduced to 204.96Mb, the compression ratio is 8.38, and the computational amount is reduced by 0.96Gflops, which facilitates the hardware deployment. Given the problem that lightweight Dense Net has no practical optimization in storage and computing after sparsing, this paper innovatively proposes a CSB compression storage method and supporting sparse convolution algorithm, which can effectively reduce the computing and storage requirements of inference network, realize the real computing acceleration and storage optimization, and overcome the hardware deployment problems. Compared with the original network, the accuracy loss is not >2 %, the number of parameters is reduced to 204.96Mb, the compression ratio is 8.38, and the computational amount is reduced by 0.96Gflops, which facilitates the hardware deployment. Given the problem that lightweight Dense Net has no practical optimization in storage and computing after sparsing, this paper innovatively proposes a CSB compression storage method and supporting sparse convolution algorithm, which can effectively reduce the computing and storage requirements of inference network, realize the real computing acceleration and storage optimization, and overcome the hardware deployment problems.
{"title":"Research on e-commerce special commodity recommendation system based on attention mechanism and Dense Net model","authors":"Daocai Wang","doi":"10.1016/j.sasc.2025.200216","DOIUrl":"10.1016/j.sasc.2025.200216","url":null,"abstract":"<div><div>This paper constructs two cross-domain recommendation models based on the perspective of user sharing and non-sharing, both of which rely on intensive convolutional networks and attention mechanisms. This research introduces lightweight Dense Net and fine-grained pruning for model optimization. Lightweight Dense Net retains the core advantages by optimizing the repeat structure while reducing redundant parameters. Compared with the original network, the accuracy loss is not >2 %, the number of parameters is reduced to 204.96Mb, the compression ratio is 8.38, and the computational amount is reduced by 0.96Gflops, which facilitates the hardware deployment. Given the problem that lightweight Dense Net has no practical optimization in storage and computing after sparsing, this paper innovatively proposes a CSB compression storage method and supporting sparse convolution algorithm, which can effectively reduce the computing and storage requirements of inference network, realize the real computing acceleration and storage optimization, and overcome the hardware deployment problems. Compared with the original network, the accuracy loss is not >2 %, the number of parameters is reduced to 204.96Mb, the compression ratio is 8.38, and the computational amount is reduced by 0.96Gflops, which facilitates the hardware deployment. Given the problem that lightweight Dense Net has no practical optimization in storage and computing after sparsing, this paper innovatively proposes a CSB compression storage method and supporting sparse convolution algorithm, which can effectively reduce the computing and storage requirements of inference network, realize the real computing acceleration and storage optimization, and overcome the hardware deployment problems.</div></div>","PeriodicalId":101205,"journal":{"name":"Systems and Soft Computing","volume":"7 ","pages":"Article 200216"},"PeriodicalIF":0.0,"publicationDate":"2025-03-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143685570","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-07DOI: 10.1016/j.sasc.2025.200214
Jie Zheng
Tourism supply chain (TSC) resilience is a measure of the TSC's response to external risks. Currently, intelligent models related to TSC resilience are basically blank. This article is based on the Collaborative Planning Forecasting and Replenishmen (CPFR) model to study the supply chain collaboration mode of smart tourism, providing a train of thought for the research of smart tourism, the purpose is to further improve the accuracy of tourism supply chain toughness assessment, and provide theoretical support for scenic spots to improve their own supply chain toughness. Simultaneously combining machine learning methods to construct a supply chain collaborative prediction model provides a new approach for collaborative prediction in the supply chain. This paper proposes a collaborative model of smart TSC based on CPFR, which not only reflects the operation process of smart TSC, but also incorporates the idea of CPFR to integrate the smart TSC into a system that can operate stably and effectively. Moreover, this paper proposes a resilience evaluation and forecasting algorithm of TSC combining complex network and genetic algorithm with genetic algorithm. In addition, this paper predicts the ability of TSC to cope with external shocks while assessing the resilience of TSC. Finally, according to the experimental research results, the model can converge after 50 iterations, and the prediction error accuracy of the test set is 5.68%, which is improved compared with the existing models The most important influencing factor in the evaluation of tourism supply chain elasticity is the tourist attractions themselves, followed by the economic environment and tourism facilities and services. Under the premise of investment level of 100, the evaluation results of the three are 33.25, 19, 19, respectively. The model proposed in this paper can realize the early forecasting of the TSC, improve the ability of the TSC to cope with risks, and promote the effective improvement of the resilience of the TSC.
{"title":"Tourism supply chain resilience assessment and optimization based on complex networks and genetic algorithms","authors":"Jie Zheng","doi":"10.1016/j.sasc.2025.200214","DOIUrl":"10.1016/j.sasc.2025.200214","url":null,"abstract":"<div><div>Tourism supply chain (TSC) resilience is a measure of the TSC's response to external risks. Currently, intelligent models related to TSC resilience are basically blank. This article is based on the Collaborative Planning Forecasting and Replenishmen (CPFR) model to study the supply chain collaboration mode of smart tourism, providing a train of thought for the research of smart tourism, the purpose is to further improve the accuracy of tourism supply chain toughness assessment, and provide theoretical support for scenic spots to improve their own supply chain toughness. Simultaneously combining machine learning methods to construct a supply chain collaborative prediction model provides a new approach for collaborative prediction in the supply chain. This paper proposes a collaborative model of smart TSC based on CPFR, which not only reflects the operation process of smart TSC, but also incorporates the idea of CPFR to integrate the smart TSC into a system that can operate stably and effectively. Moreover, this paper proposes a resilience evaluation and forecasting algorithm of TSC combining complex network and genetic algorithm with genetic algorithm. In addition, this paper predicts the ability of TSC to cope with external shocks while assessing the resilience of TSC. Finally, according to the experimental research results, the model can converge after 50 iterations, and the prediction error accuracy of the test set is 5.68%, which is improved compared with the existing models The most important influencing factor in the evaluation of tourism supply chain elasticity is the tourist attractions themselves, followed by the economic environment and tourism facilities and services. Under the premise of investment level of 100, the evaluation results of the three are 33.25, 19, 19, respectively. The model proposed in this paper can realize the early forecasting of the TSC, improve the ability of the TSC to cope with risks, and promote the effective improvement of the resilience of the TSC.</div></div>","PeriodicalId":101205,"journal":{"name":"Systems and Soft Computing","volume":"7 ","pages":"Article 200214"},"PeriodicalIF":0.0,"publicationDate":"2025-03-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143562667","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号