Pub Date : 2024-07-18DOI: 10.1109/TCSS.2024.3410391
Jianghong Zhu;Zhenwen Zhang;Zhihua Guo;Zepeng Li
Anxiety disorder is a common mental disorder that has received increasing attention due to its high incidence, comorbidity, and recurrence. In recent years, with the rapid development of information technology, social media platforms have become a crucial source of data for studying anxiety disorders. Existing studies on anxiety disorders have focused on utilizing user-generated contents to study correlations with disorders or identify disorders. However, these studies overlook the emotional information in social media posts, restraining the effective capture of users’ emotions or mental states when posting. This article focuses on the sentiment polarity of anxiety-related posts on a Chinese social media and designs sentiment classification models via fuzing linguistic and semantic features of the posts. First, we extract the linguistic features from posts based on the simplified Chinese–Linguistic inquiry and word count (SC-LIWC) dictionary, and propose a novel recursive feature selection algorithm to reserve important linguistic features. Second, we propose a TextCNN-based model to study the deep semantic features of posts and fuze their linguistic features to obtain a better representation. Finally, to conduct anxiety analysis on Chinese social media, we construct a postlevel sentiment analysis dataset based on anxiety-related posts on Sina Weibo. The experimental results indicate that our proposed fusion models exhibit better performance in the task of identifying the sentiment polarity of anxiety-related posts on Chinese social media.
{"title":"Sentiment Classification of Anxiety-Related Texts in Social Media via Fuzing Linguistic and Semantic Features","authors":"Jianghong Zhu;Zhenwen Zhang;Zhihua Guo;Zepeng Li","doi":"10.1109/TCSS.2024.3410391","DOIUrl":"https://doi.org/10.1109/TCSS.2024.3410391","url":null,"abstract":"Anxiety disorder is a common mental disorder that has received increasing attention due to its high incidence, comorbidity, and recurrence. In recent years, with the rapid development of information technology, social media platforms have become a crucial source of data for studying anxiety disorders. Existing studies on anxiety disorders have focused on utilizing user-generated contents to study correlations with disorders or identify disorders. However, these studies overlook the emotional information in social media posts, restraining the effective capture of users’ emotions or mental states when posting. This article focuses on the sentiment polarity of anxiety-related posts on a Chinese social media and designs sentiment classification models via fuzing linguistic and semantic features of the posts. First, we extract the linguistic features from posts based on the simplified Chinese–Linguistic inquiry and word count (SC-LIWC) dictionary, and propose a novel recursive feature selection algorithm to reserve important linguistic features. Second, we propose a TextCNN-based model to study the deep semantic features of posts and fuze their linguistic features to obtain a better representation. Finally, to conduct anxiety analysis on Chinese social media, we construct a postlevel sentiment analysis dataset based on anxiety-related posts on Sina Weibo. The experimental results indicate that our proposed fusion models exhibit better performance in the task of identifying the sentiment polarity of anxiety-related posts on Chinese social media.","PeriodicalId":13044,"journal":{"name":"IEEE Transactions on Computational Social Systems","volume":"11 5","pages":"6819-6829"},"PeriodicalIF":4.5,"publicationDate":"2024-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142368279","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-07-18DOI: 10.1109/TCSS.2024.3418625
Emiliano Alvarez;Marcelo Álvez;Juan Gabriel Brida
In this article, we apply an agent-based stock-flow consistent model (AB-SFC) to analyze economic growth differences when establishing different types of taxes on personal income: proportional and progressive. We use an income tax design that distinguishes between two sections of income. We contribute to the prominent literature on macro agent-based models by providing an unexplored feature in the income tax scheme. Our main findings are that this tax design seems to offset the inequality through tax exemption for low-income households but seems to have a limited impact on inequality generated between middle and high-income households. Notably, we did not find evidence of a deterioration in economic growth in the presence of a progressive income tax instead of a proportional one. Therefore, this article proposes a scenario where changing the tax scheme reduces inequality without hampering growth. This result has important implications for policy.
{"title":"Progressive Income Tax and Its Emerging Growth Effects: A Complex System Approach","authors":"Emiliano Alvarez;Marcelo Álvez;Juan Gabriel Brida","doi":"10.1109/TCSS.2024.3418625","DOIUrl":"https://doi.org/10.1109/TCSS.2024.3418625","url":null,"abstract":"In this article, we apply an agent-based stock-flow consistent model (AB-SFC) to analyze economic growth differences when establishing different types of taxes on personal income: proportional and progressive. We use an income tax design that distinguishes between two sections of income. We contribute to the prominent literature on macro agent-based models by providing an unexplored feature in the income tax scheme. Our main findings are that this tax design seems to offset the inequality through tax exemption for low-income households but seems to have a limited impact on inequality generated between middle and high-income households. Notably, we did not find evidence of a deterioration in economic growth in the presence of a progressive income tax instead of a proportional one. Therefore, this article proposes a scenario where changing the tax scheme reduces inequality without hampering growth. This result has important implications for policy.","PeriodicalId":13044,"journal":{"name":"IEEE Transactions on Computational Social Systems","volume":"11 5","pages":"6605-6622"},"PeriodicalIF":4.5,"publicationDate":"2024-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142368225","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-07-18DOI: 10.1109/TCSS.2024.3421672
Priyanshu Priya;Mauajama Firdaus;Asif Ekbal
Politeness is key to successful conversations. It depicts the behavior that is socially valued and is often accompanied by emotions. Previously, researchers have focused on detecting politeness in goal-oriented conversations in high-resource English language. The existing studies do not focus on identifying politeness in a resource-scared Indian languages such as Hindi, primarily due to the lack of labeled data. To overcome this limitation, in this article, we propose a novel emotion-aware contrastive learning (CL) method for zero-shot cross-lingual politeness identification ( XeroPol ) task in dialogues. We introduce ContrastiveAligner , a CL-based alignment method for zero-shot cross-lingual transfer. ContrastiveAligner employs translated data and pushes the model to generate similar utterance embeddings for different languages. As politeness and emotion are interrelated, hence, as the conversation progresses, the variation in emotions tends to pose challenges in identifying politeness in dialogues. Thus, in this work, we also design an auxiliary emotion-aware CL objective using sentiment information, namely the EmoSenti objective , which is expected to implicitly model the emotion change across utterances and help in the primary task of politeness identification. Experiments on MultiDoGo and EmoWOZ datasets demonstrate that the proposed approach significantly outperforms the baselines. Further analysis such as human evaluation on the EmoInHindi dataset validates the efficacy of the entire approach.
{"title":"XeroPol: Emotion-Aware Contrastive Learning for Zero-Shot Cross-Lingual Politeness Identification in Dialogues","authors":"Priyanshu Priya;Mauajama Firdaus;Asif Ekbal","doi":"10.1109/TCSS.2024.3421672","DOIUrl":"https://doi.org/10.1109/TCSS.2024.3421672","url":null,"abstract":"Politeness is key to successful conversations. It depicts the behavior that is socially valued and is often accompanied by emotions. Previously, researchers have focused on detecting politeness in goal-oriented conversations in high-resource English language. The existing studies do not focus on identifying politeness in a resource-scared Indian languages such as Hindi, primarily due to the lack of labeled data. To overcome this limitation, in this article, we propose a novel emotion-aware contrastive learning (CL) method for zero-shot cross-lingual politeness identification (\u0000<italic>XeroPol</i>\u0000) task in dialogues. We introduce \u0000<italic>ContrastiveAligner</i>\u0000, a CL-based alignment method for zero-shot cross-lingual transfer. \u0000<italic>ContrastiveAligner</i>\u0000 employs translated data and pushes the model to generate similar utterance embeddings for different languages. As politeness and emotion are interrelated, hence, as the conversation progresses, the variation in emotions tends to pose challenges in identifying politeness in dialogues. Thus, in this work, we also design an auxiliary emotion-aware CL objective using sentiment information, namely the \u0000<italic>EmoSenti objective</i>\u0000, which is expected to implicitly model the emotion change across utterances and help in the primary task of politeness identification. Experiments on MultiDoGo and EmoWOZ datasets demonstrate that the proposed approach significantly outperforms the baselines. Further analysis such as human evaluation on the EmoInHindi dataset validates the efficacy of the entire approach.","PeriodicalId":13044,"journal":{"name":"IEEE Transactions on Computational Social Systems","volume":"11 5","pages":"6662-6671"},"PeriodicalIF":4.5,"publicationDate":"2024-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142368394","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-07-16DOI: 10.1109/TCSS.2024.3409833
Litong Fan;Dengxiu Yu;Zhen Wang
This article presents a framework for exploring optimal evolutionary strategies in continuous-action social dilemma games with a hierarchical structure comprising a leader and multifollowers. Previous studies in game theory have frequently overlooked the hierarchical structure among individuals, assuming that decisions are made simultaneously. Here, we propose a hierarchical structure for continuous action games that involves a leader and followers to enhance cooperation. The optimal evolutionary strategy for the leader is to guide the followers’ actions to maximize overall benefits by exerting minimal control, while the followers aim to maximize their payoff by making minimal changes to their strategies. We establish the coupled Hamilton–Jacobi–Bellman (HJB) equations to find the optimal evolutionary strategy. To address the complexity of asymmetric roles arising from the leader-follower structure, we introduce an integral reinforcement learning (RL) algorithm known as two-level heuristic dynamic programming (HDP)-based value iteration (VI). The implementation of the algorithm utilizes neural networks (NNs) to approximate the value functions. Moreover, the convergence of the proposed algorithm is demonstrated. Additionally, three social dilemma models are presented to validate the efficacy of the proposed algorithm.
{"title":"Integral-Reinforcement-Learning-Based Hierarchical Optimal Evolutionary Strategy for Continuous Action Social Dilemma Games","authors":"Litong Fan;Dengxiu Yu;Zhen Wang","doi":"10.1109/TCSS.2024.3409833","DOIUrl":"https://doi.org/10.1109/TCSS.2024.3409833","url":null,"abstract":"This article presents a framework for exploring optimal evolutionary strategies in continuous-action social dilemma games with a hierarchical structure comprising a leader and multifollowers. Previous studies in game theory have frequently overlooked the hierarchical structure among individuals, assuming that decisions are made simultaneously. Here, we propose a hierarchical structure for continuous action games that involves a leader and followers to enhance cooperation. The optimal evolutionary strategy for the leader is to guide the followers’ actions to maximize overall benefits by exerting minimal control, while the followers aim to maximize their payoff by making minimal changes to their strategies. We establish the coupled Hamilton–Jacobi–Bellman (HJB) equations to find the optimal evolutionary strategy. To address the complexity of asymmetric roles arising from the leader-follower structure, we introduce an integral reinforcement learning (RL) algorithm known as two-level heuristic dynamic programming (HDP)-based value iteration (VI). The implementation of the algorithm utilizes neural networks (NNs) to approximate the value functions. Moreover, the convergence of the proposed algorithm is demonstrated. Additionally, three social dilemma models are presented to validate the efficacy of the proposed algorithm.","PeriodicalId":13044,"journal":{"name":"IEEE Transactions on Computational Social Systems","volume":"11 5","pages":"6807-6818"},"PeriodicalIF":4.5,"publicationDate":"2024-07-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142368337","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Organized misinformation campaigns on Twitter continue to proliferate, even as the platform acknowledges such activities through its transparency center. These deceptive initiatives significantly impact vital societal issues, including climate change, thus spurring research aimed at pinpointing and intercepting these malicious actors. Present-day algorithms for detecting bots harness an array of data drawn from user profiles, tweets, and network configurations, delivering commendable outcomes. Yet, these strategies mainly concentrate on postincident identification of malevolent users, hinging on static training datasets that categorize individuals based on historical activities. Diverging from this approach, we advocate for a forward-thinking methodology, which utilizes user data to foresee and mitigate potential threats before their realization, thereby cultivating more secure, equitable, and unbiased online communities. To this end, our proposed technique forecasts malevolent activities by tracing the projected trajectories of user embeddings before any malevolent action materializes. For validation, we employed a dynamic directed multigraph paradigm to chronicle the evolving engagements between Twitter users. When juxtaposed against the identical dataset, our technique eclipses contemporary methodologies by an impressive 40.66% in F score (F1 score) in the anticipatory identification of harmful users. Furthermore, we undertook a model evaluation exercise to gauge the efficiency of distinct system elements.
推特上有组织的虚假信息传播活动继续泛滥，即使该平台通过其透明度中心承认此类活动。这些欺骗性活动对包括气候变化在内的重要社会问题产生了重大影响，从而推动了旨在精确定位和拦截这些恶意行为者的研究。目前用于检测机器人的算法利用了从用户配置文件、推文和网络配置中提取的一系列数据，取得了值得称道的成果。然而，这些策略主要集中于事后识别恶意用户，依赖于根据历史活动对个人进行分类的静态训练数据集。与这种方法不同的是，我们主张采用一种前瞻性方法，利用用户数据在潜在威胁发生之前对其进行预测和缓解，从而培养出更加安全、公平和公正的网络社区。为此，我们提出的技术可以在任何恶意行为发生之前，通过追踪用户嵌入的预测轨迹来预测恶意活动。为了进行验证，我们采用了动态有向多图范例来记录 Twitter 用户之间不断发展的互动。在与相同的数据集进行对比时，我们的技术在有害用户的预测性识别方面以令人印象深刻的 40.66% 的 F 分数（F1 分数）超越了当代的方法。此外，我们还进行了模型评估，以衡量不同系统元素的效率。
{"title":"Early Detection and Prevention of Malicious User Behavior on Twitter Using Deep Learning Techniques","authors":"Rubén Sánchez-Corcuera;Arkaitz Zubiaga;Aitor Almeida","doi":"10.1109/TCSS.2024.3419171","DOIUrl":"https://doi.org/10.1109/TCSS.2024.3419171","url":null,"abstract":"Organized misinformation campaigns on Twitter continue to proliferate, even as the platform acknowledges such activities through its transparency center. These deceptive initiatives significantly impact vital societal issues, including climate change, thus spurring research aimed at pinpointing and intercepting these malicious actors. Present-day algorithms for detecting bots harness an array of data drawn from user profiles, tweets, and network configurations, delivering commendable outcomes. Yet, these strategies mainly concentrate on postincident identification of malevolent users, hinging on static training datasets that categorize individuals based on historical activities. Diverging from this approach, we advocate for a forward-thinking methodology, which utilizes user data to foresee and mitigate potential threats before their realization, thereby cultivating more secure, equitable, and unbiased online communities. To this end, our proposed technique forecasts malevolent activities by tracing the projected trajectories of user embeddings before any malevolent action materializes. For validation, we employed a dynamic directed multigraph paradigm to chronicle the evolving engagements between Twitter users. When juxtaposed against the identical dataset, our technique eclipses contemporary methodologies by an impressive 40.66% in F score (F1 score) in the anticipatory identification of harmful users. Furthermore, we undertook a model evaluation exercise to gauge the efficiency of distinct system elements.","PeriodicalId":13044,"journal":{"name":"IEEE Transactions on Computational Social Systems","volume":"11 5","pages":"6649-6661"},"PeriodicalIF":4.5,"publicationDate":"2024-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10597373","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142368534","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Quasi group role assignment (QGRA) presents a novel social computing model designed to address the burgeoning domain of self-service spatiotemporal crowdsourcing (SSC), specifically for tackling the photographing to make money problem (PMMP). Nevertheless, the application of QGRA in practical scenarios encounters a significant bottleneck. QGRA provides optimal assignment strategies under conditions where both the number of crowdsourced tasks and workers remain stable. However, real-world crowdsourcing applications may necessitate the phased integration of new tasks. With the rapid increase in the number of tasks, a set of residual tasks inevitably exists that are difficult to complete. To maximize the completion of crowdsourced tasks, workers may be assigned low-yield or even unprofitable tasks. Given the reluctance of crowdsourcing workers to be overstretched for these tasks, along with the inherent characteristics of self-service crowdsourcing tasks, this can lead to the failure of the assignment scheme. To tackle the identified challenges, this article proposes the QGRA with agent satisfaction (QGRAAS) method. Initially, it sheds light on a creative satisfaction filtering algorithm (SFA), which is engineered to perform optimal task assignments while actively optimizing the profitability of crowdsourcing workers. This approach ensures the satisfaction of workers, thereby fostering their loyalty to the platform. Concurrently, in response to the phased changes in the crowdsourcing environment, this article incorporates the concept of bonus incentives. This aids decision-makers in achieving a tradeoff between the operational costs and task completion rates. The robustness and practicality of the proposed solutions are confirmed through simulation experiments.
{"title":"Quasi Group Role Assignment With Agent Satisfaction in Self-Service Spatiotemporal Crowdsourcing","authors":"Qian Jiang;Dongning Liu;Haibin Zhu;Baoying Huang;Naiqi Wu;Yan Qiao","doi":"10.1109/TCSS.2024.3417959","DOIUrl":"https://doi.org/10.1109/TCSS.2024.3417959","url":null,"abstract":"Quasi group role assignment (QGRA) presents a novel social computing model designed to address the burgeoning domain of self-service spatiotemporal crowdsourcing (SSC), specifically for tackling the photographing to make money problem (PMMP). Nevertheless, the application of QGRA in practical scenarios encounters a significant bottleneck. QGRA provides optimal assignment strategies under conditions where both the number of crowdsourced tasks and workers remain stable. However, real-world crowdsourcing applications may necessitate the phased integration of new tasks. With the rapid increase in the number of tasks, a set of residual tasks inevitably exists that are difficult to complete. To maximize the completion of crowdsourced tasks, workers may be assigned low-yield or even unprofitable tasks. Given the reluctance of crowdsourcing workers to be overstretched for these tasks, along with the inherent characteristics of self-service crowdsourcing tasks, this can lead to the failure of the assignment scheme. To tackle the identified challenges, this article proposes the QGRA with agent satisfaction (QGRAAS) method. Initially, it sheds light on a creative satisfaction filtering algorithm (SFA), which is engineered to perform optimal task assignments while actively optimizing the profitability of crowdsourcing workers. This approach ensures the satisfaction of workers, thereby fostering their loyalty to the platform. Concurrently, in response to the phased changes in the crowdsourcing environment, this article incorporates the concept of bonus incentives. This aids decision-makers in achieving a tradeoff between the operational costs and task completion rates. The robustness and practicality of the proposed solutions are confirmed through simulation experiments.","PeriodicalId":13044,"journal":{"name":"IEEE Transactions on Computational Social Systems","volume":"11 5","pages":"7002-7019"},"PeriodicalIF":4.5,"publicationDate":"2024-07-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142368540","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The ability to anticipate global technical expertise and capability evolution trends is essential for national and global security, especially in safety-critical domains such as nuclear nonproliferation (NN) and rapidly emerging fields like artificial intelligence (AI). In this work, we extend traditional statistical relational learning approaches (e.g., link prediction in collaboration networks) and formulate a problem of anticipating technical expertise and capability evolution using dynamic heterogeneous graph representations. We develop novel capabilities to forecast collaboration patterns, authorship behavior, and technical capability evolution at different granularities (e.g., scientist and institution levels) in two distinct research fields. We implement a dynamic graph transformer (DGT) neural architecture, which pushes the state-of-the-art graph neural network models by: 1) forecasting heterogeneous (rather than homogeneous) nodes and edges; and 2) relying on both discrete- and continuous-time inputs. We demonstrate that our DGT models predict collaboration, partnership, and expertise patterns with 0.26, 0.73, and 0.53 mean reciprocal rank values for AI and 0.48, 0.93, and 0.22 for NN domains. DGT model performance exceeds the best-performing static graph baseline models by 30%–80% across AI and NN domains. Our findings demonstrate that DGT models boost inductive task performance when previously unseen nodes appear in the test data for the domains with emerging collaboration patterns (e.g., AI). Specifically, models accurately predict which established scientists will collaborate with early career scientists and vice versa in the AI domain.
{"title":"Anticipating Technical Expertise and Capability Evolution in Research Communities Using Dynamic Graph Transformers","authors":"Sameera Horawalavithana;Ellyn Ayton;Anastasiya Usenko;Robin Cosbey;Svitlana Volkova","doi":"10.1109/TCSS.2024.3416837","DOIUrl":"https://doi.org/10.1109/TCSS.2024.3416837","url":null,"abstract":"The ability to anticipate global technical expertise and capability evolution trends is essential for national and global security, especially in safety-critical domains such as nuclear nonproliferation (NN) and rapidly emerging fields like artificial intelligence (AI). In this work, we extend traditional statistical relational learning approaches (e.g., link prediction in collaboration networks) and formulate a problem of anticipating technical expertise and capability evolution using dynamic heterogeneous graph representations. We develop novel capabilities to forecast collaboration patterns, authorship behavior, and technical capability evolution at different granularities (e.g., scientist and institution levels) in two distinct research fields. We implement a dynamic graph transformer (DGT) neural architecture, which pushes the state-of-the-art graph neural network models by: 1) forecasting heterogeneous (rather than homogeneous) nodes and edges; and 2) relying on both discrete- and continuous-time inputs. We demonstrate that our DGT models predict collaboration, partnership, and expertise patterns with 0.26, 0.73, and 0.53 mean reciprocal rank values for AI and 0.48, 0.93, and 0.22 for NN domains. DGT model performance exceeds the best-performing static graph baseline models by 30%–80% across AI and NN domains. Our findings demonstrate that DGT models boost inductive task performance when previously unseen nodes appear in the test data for the domains with emerging collaboration patterns (e.g., AI). Specifically, models accurately predict which established scientists will collaborate with early career scientists and vice versa in the AI domain.","PeriodicalId":13044,"journal":{"name":"IEEE Transactions on Computational Social Systems","volume":"11 5","pages":"6982-7001"},"PeriodicalIF":4.5,"publicationDate":"2024-07-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142368361","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-07-11DOI: 10.1109/TCSS.2024.3406512
Sandip Kumar Burnwal;Pragati Sinha;Bhumika;Jayant Vyas;Debasis Das
Humans relocate for a variety of reasons, including employment, study, tourism, family, and health. However, in COVID-19, the government imposed restrictions such as lockdowns, travel bans, and quarantine regulations, preventing many people from traveling for work, study, or leisure; thus, human mobility exhibits distinct patterns than ordinary movements. In this article, we analyze the effect of COVID-19 on interregion human mobility using curated Twitter data and propose a framework named Co-Move for human mobility prediction. There were three challenges in predicting mobility: 1) heterogenous data; 2) short and long-term periodic patterns; and 3) complex intercorrelation. To address these challenges, the framework comprises parallel multiscale convolution and long short-term memory components. Extensive experiments on real-life mobility datasets show the mean square error (MSE) of 0.0179, RMSE of 0.129, mean absolute error (MAE) of 0.1075, and outperform baseline models.
{"title":"Co-Move: COVID-19 and Inter-Region Human Mobility Analysis and Prediction","authors":"Sandip Kumar Burnwal;Pragati Sinha;Bhumika;Jayant Vyas;Debasis Das","doi":"10.1109/TCSS.2024.3406512","DOIUrl":"https://doi.org/10.1109/TCSS.2024.3406512","url":null,"abstract":"Humans relocate for a variety of reasons, including employment, study, tourism, family, and health. However, in COVID-19, the government imposed restrictions such as lockdowns, travel bans, and quarantine regulations, preventing many people from traveling for work, study, or leisure; thus, human mobility exhibits distinct patterns than ordinary movements. In this article, we analyze the effect of COVID-19 on interregion human mobility using curated Twitter data and propose a framework named \u0000<italic>Co-Move</i>\u0000 for human mobility prediction. There were three challenges in predicting mobility: 1) heterogenous data; 2) short and long-term periodic patterns; and 3) complex intercorrelation. To address these challenges, the framework comprises parallel multiscale convolution and long short-term memory components. Extensive experiments on real-life mobility datasets show the mean square error (MSE) of 0.0179, RMSE of 0.129, mean absolute error (MAE) of 0.1075, and outperform baseline models.","PeriodicalId":13044,"journal":{"name":"IEEE Transactions on Computational Social Systems","volume":"11 5","pages":"6843-6853"},"PeriodicalIF":4.5,"publicationDate":"2024-07-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142377009","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-07-08DOI: 10.1109/TCSS.2024.3417184
Cheng Ma;Brendan Cross;Gyorgy Korniss;Boleslaw K. Szymanski
The naming game (NG) is a classic model for studying the emergence and evolution of language within a population. In this article, we extend the traditional NG model to encompass multiple committed opinions and investigate the system dynamics on the complete graph with an arbitrarily large population and random networks of finite size. For the fully connected complete graph, the homogeneous mixing condition enables us to use mean-field theory to analyze the opinion evolution of the system. However, when the number of opinions increases, the number of variables describing the system grows exponentially. To mitigate this, we focus on a special scenario where the largest group of committed agents competes with a motley of committed groups, each of which is smaller than the largest one, while initially, most of uncommitted agents hold one unique opinion. This scenario is chosen for its recurrence in diverse societies and its potential for complexity reduction by unifying agents from smaller committed groups into one category. Our investigation reveals that when the size of the largest committed group reaches the critical threshold, most of uncommitted agents change their beliefs to this opinion, triggering a phase transition. Further, we derive the general formula for the multiopinion evolution using a recursive approach, enabling investigation into any scenario. Finally, we employ agent-based simulations to reveal the opinion evolution and dominance transition in random graphs. Our results provide insights into the conditions under which the dominant opinion emerges in a population and the factors that influence these conditions.
命名游戏（NG）是研究群体中语言出现和演变的经典模型。在本文中，我们将传统的 NG 模型扩展到包含多种承诺意见，并研究了具有任意大群体和有限大小随机网络的完整图上的系统动态。对于全连接的完整图，同质混合条件使我们能够使用均值场理论来分析系统的意见演化。然而，当意见数量增加时，描述系统的变量数量会呈指数增长。为了缓解这一问题，我们重点研究了一种特殊情况，即最大的承诺代理群体与多个承诺代理群体竞争，每个承诺代理群体都比最大的承诺代理群体小，而最初，大多数未承诺代理群体都持有一种独特的意见。之所以选择这种情况，是因为它在多样化的社会中经常出现，而且通过将较小的承诺群体中的代理人统一到一个类别中，有可能降低复杂性。我们的研究发现，当最大的承诺群体的规模达到临界阈值时，大多数未承诺的代理会改变他们的信念，转而持有这种观点，从而引发阶段性转变。此外，我们还利用递归方法推导出了多意见演变的一般公式，从而可以对任何情况进行研究。最后，我们采用基于代理的模拟来揭示随机图中的观点演变和优势转换。我们的研究结果让我们深入了解了主导观点在群体中出现的条件以及影响这些条件的因素。
{"title":"Divide-and-Conquer Policy in the Naming Game","authors":"Cheng Ma;Brendan Cross;Gyorgy Korniss;Boleslaw K. Szymanski","doi":"10.1109/TCSS.2024.3417184","DOIUrl":"https://doi.org/10.1109/TCSS.2024.3417184","url":null,"abstract":"The naming game (NG) is a classic model for studying the emergence and evolution of language within a population. In this article, we extend the traditional NG model to encompass multiple committed opinions and investigate the system dynamics on the complete graph with an arbitrarily large population and random networks of finite size. For the fully connected complete graph, the homogeneous mixing condition enables us to use mean-field theory to analyze the opinion evolution of the system. However, when the number of opinions increases, the number of variables describing the system grows exponentially. To mitigate this, we focus on a special scenario where the largest group of committed agents competes with a motley of committed groups, each of which is smaller than the largest one, while initially, most of uncommitted agents hold one unique opinion. This scenario is chosen for its recurrence in diverse societies and its potential for complexity reduction by unifying agents from smaller committed groups into one category. Our investigation reveals that when the size of the largest committed group reaches the critical threshold, most of uncommitted agents change their beliefs to this opinion, triggering a phase transition. Further, we derive the general formula for the multiopinion evolution using a recursive approach, enabling investigation into any scenario. Finally, we employ agent-based simulations to reveal the opinion evolution and dominance transition in random graphs. Our results provide insights into the conditions under which the dominant opinion emerges in a population and the factors that influence these conditions.","PeriodicalId":13044,"journal":{"name":"IEEE Transactions on Computational Social Systems","volume":"11 5","pages":"6911-6924"},"PeriodicalIF":4.5,"publicationDate":"2024-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142368508","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-07-04DOI: 10.1109/TCSS.2024.3406501
Gabriel Stechschulte;Malte Wintner;Matthias Hemmje;Jürg Schwarz;Suzanne Lischer;Michael Kaufmann
This study involves a comprehensive analysis of an anonymized dataset provided by a Swiss online casino that adds to the identification of reliable early indicators for problematic online gambling. Targeting gambling addiction prevention, our objective was to model and evaluate behavioral characteristics that signal early stages of problem gambling. We scrutinized player behaviors against a list of gamblers previously excluded for problematic gambling, using this as our target variable. Our approach combined traditional gambling risk indicators, as outlined in the existing literature, with innovative exploratory feature engineering and feature selection. This involved computing moving aggregates over specific periods to capture nuanced gambling patterns. All features were evaluated by assessing mutual information with the target variable as well as the collinearity of each pairwise combination of features. Based on our data analysis, we found that the total losses in the previous seven days, total deposits in the previous 15 days, total duration played in the previous seven days, stakes (amount bet per game) over the previous seven days, and making a deposit 12 h after a loss (chasing) were the most informative and independent risk indicators. To assess the accuracy of these indicators for early detection of problematic gambling and accordingly for responsible gambling interventions, we combined them in a linear regression model and compared its performance with the casino's currently used model. We found that a binary decision model based on a linear combination of these indicators provided better recall, greater precision, and more timely decisions than the benchmark.
{"title":"In-Database Feature Extraction to Improve Early Detection of Problematic Online Gambling Behavior","authors":"Gabriel Stechschulte;Malte Wintner;Matthias Hemmje;Jürg Schwarz;Suzanne Lischer;Michael Kaufmann","doi":"10.1109/TCSS.2024.3406501","DOIUrl":"https://doi.org/10.1109/TCSS.2024.3406501","url":null,"abstract":"This study involves a comprehensive analysis of an anonymized dataset provided by a Swiss online casino that adds to the identification of reliable early indicators for problematic online gambling. Targeting gambling addiction prevention, our objective was to model and evaluate behavioral characteristics that signal early stages of problem gambling. We scrutinized player behaviors against a list of gamblers previously excluded for problematic gambling, using this as our target variable. Our approach combined traditional gambling risk indicators, as outlined in the existing literature, with innovative exploratory feature engineering and feature selection. This involved computing moving aggregates over specific periods to capture nuanced gambling patterns. All features were evaluated by assessing mutual information with the target variable as well as the collinearity of each pairwise combination of features. Based on our data analysis, we found that the total losses in the previous seven days, total deposits in the previous 15 days, total duration played in the previous seven days, stakes (amount bet per game) over the previous seven days, and making a deposit 12 h after a loss (chasing) were the most informative and independent risk indicators. To assess the accuracy of these indicators for early detection of problematic gambling and accordingly for responsible gambling interventions, we combined them in a linear regression model and compared its performance with the casino's currently used model. We found that a binary decision model based on a linear combination of these indicators provided better recall, greater precision, and more timely decisions than the benchmark.","PeriodicalId":13044,"journal":{"name":"IEEE Transactions on Computational Social Systems","volume":"11 5","pages":"6868-6881"},"PeriodicalIF":4.5,"publicationDate":"2024-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142368338","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：