The current prevalence of conspiracy theories on the internet is a significant issue, tackled by many computational approaches. However, these approaches fail to recognize the relevance of distinguishing between texts which contain a conspiracy theory and texts which are simply critical and oppose mainstream narratives. Furthermore, little attention is usually paid to the role of inter‐group conflict in oppositional narratives. We contribute by proposing a novel topic‐agnostic annotation scheme that differentiates between conspiracies and critical texts, and that defines span‐level categories of inter‐group conflict. We also contribute with the multilingual XAI‐DisInfodemics corpus (English and Spanish), which contains a high‐quality annotation of Telegram messages related to COVID‐19 (5000 messages per language). We also demonstrate the feasibility of an NLP‐based automatization by performing a range of experiments that yield strong baseline solutions. Finally, we perform an analysis which demonstrates that the promotion of intergroup conflict and the presence of violence and anger are key aspects to distinguish between the two types of oppositional narratives, that is, conspiracy versus critical.
{"title":"What distinguishes conspiracy from critical narratives? A computational analysis of oppositional discourse","authors":"Damir Korenčić, Berta Chulvi, Xavier Bonet Casals, Alejandro Toselli, Mariona Taulé, Paolo Rosso","doi":"10.1111/exsy.13671","DOIUrl":"https://doi.org/10.1111/exsy.13671","url":null,"abstract":"The current prevalence of conspiracy theories on the internet is a significant issue, tackled by many computational approaches. However, these approaches fail to recognize the relevance of distinguishing between texts which contain a conspiracy theory and texts which are simply critical and oppose mainstream narratives. Furthermore, little attention is usually paid to the role of inter‐group conflict in oppositional narratives. We contribute by proposing a novel topic‐agnostic annotation scheme that differentiates between conspiracies and critical texts, and that defines span‐level categories of inter‐group conflict. We also contribute with the multilingual XAI‐DisInfodemics corpus (English and Spanish), which contains a high‐quality annotation of Telegram messages related to COVID‐19 (5000 messages per language). We also demonstrate the feasibility of an NLP‐based automatization by performing a range of experiments that yield strong baseline solutions. Finally, we perform an analysis which demonstrates that the promotion of intergroup conflict and the presence of violence and anger are key aspects to distinguish between the two types of oppositional narratives, that is, conspiracy versus critical.","PeriodicalId":51053,"journal":{"name":"Expert Systems","volume":null,"pages":null},"PeriodicalIF":3.3,"publicationDate":"2024-07-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141586682","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"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 forecasting of steel futures prices is important for the steel futures market, even for the steel industry. We propose a decomposition ensemble model that incorporates the Ensemble Empirical Mode Decomposition (EEMD), Long Short‐Term Memory (LSTM), Support Vector Regression (SVR), and Back Propagation (BP) neural network to forecast steel futures prices. The forecasting procedures are as follows: (1) The price data are initially decomposed into several relatively independent Intrinsic Mode Functions (IMFs) and a residue using EEMD. (2) The IMFs are then reconstructed as components representing short‐term, medium‐term, and long‐term frequencies via fine‐to‐coarse. (3) LSTM, SVR, and BP neural network are utilized to forecast the short‐term, medium‐term, and long‐term reconstructed components, respectively. (4) The prediction results for each component are simply added to the final prediction results. The accuracy of the proposed model is compared with several benchmark models by experiments and evaluated by some prediction evaluation indexes. The experimental results show that our model outperforms other models in terms of forecast accuracy, confirming its strong predictive capabilities. This study provides some suggestions for investment and decision making by participants in the steel futures market. It may promote the smooth operation of the steel futures market and shed some light on the operation of the steel industry.
钢材期货价格预测对于钢材期货市场乃至钢铁行业都非常重要。我们提出了一种分解集合模型,该模型融合了集合经验模式分解(EEMD)、长短期记忆(LSTM)、支持向量回归(SVR)和反向传播(BP)神经网络,用于预测钢材期货价格。预测程序如下(1) 首先使用 EEMD 将价格数据分解为几个相对独立的本征模式函数(IMF)和一个残差。(2) 然后通过从细到粗的方法将 IMF 重构为代表短期、中期和长期频率的成分。(3) 利用 LSTM、SVR 和 BP 神经网络分别预测重建的短期、中期和长期分量。(4) 将各分量的预测结果简单相加,得出最终预测结果。通过实验将所提出模型的准确性与几个基准模型进行比较,并通过一些预测评价指标进行评估。实验结果表明,我们的模型在预测准确率方面优于其他模型,证实了其强大的预测能力。本研究为钢铁期货市场参与者的投资和决策提供了一些建议。它可以促进钢材期货市场的平稳运行,并对钢铁行业的运行起到一定的启示作用。
{"title":"An EEMD‐LSTM, SVR, and BP decomposition ensemble model for steel future prices forecasting","authors":"Sen Wu, Wei Wang, Yanan Song, Shuaiqi Liu","doi":"10.1111/exsy.13672","DOIUrl":"https://doi.org/10.1111/exsy.13672","url":null,"abstract":"The forecasting of steel futures prices is important for the steel futures market, even for the steel industry. We propose a decomposition ensemble model that incorporates the Ensemble Empirical Mode Decomposition (EEMD), Long Short‐Term Memory (LSTM), Support Vector Regression (SVR), and Back Propagation (BP) neural network to forecast steel futures prices. The forecasting procedures are as follows: (1) The price data are initially decomposed into several relatively independent Intrinsic Mode Functions (IMFs) and a residue using EEMD. (2) The IMFs are then reconstructed as components representing short‐term, medium‐term, and long‐term frequencies via fine‐to‐coarse. (3) LSTM, SVR, and BP neural network are utilized to forecast the short‐term, medium‐term, and long‐term reconstructed components, respectively. (4) The prediction results for each component are simply added to the final prediction results. The accuracy of the proposed model is compared with several benchmark models by experiments and evaluated by some prediction evaluation indexes. The experimental results show that our model outperforms other models in terms of forecast accuracy, confirming its strong predictive capabilities. This study provides some suggestions for investment and decision making by participants in the steel futures market. It may promote the smooth operation of the steel futures market and shed some light on the operation of the steel industry.","PeriodicalId":51053,"journal":{"name":"Expert Systems","volume":null,"pages":null},"PeriodicalIF":3.3,"publicationDate":"2024-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141570209","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"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 people in the world rely on social media for gathering news, and it is mainly because of the development of technology. The approaches employed in natural language processing are still deficient in judgement factors, and these techniques frequently rely upon political or social circumstances. Numerous low‐level communities in the area are curious after experiencing the negative effects caused by the spread of false information in different sectors. Low‐resource languages are still distracted, because these techniques are extensively employed in the English language. This work aims to provide an analysis of regional language fake news and develop a referral system with advanced techniques to identify fake news in Hindi and Tamil. This proposed model includes (a) Regional Language Text Collection; (b) Text preprocessing; (c) Feature Extraction; (d) Weighted Stacked Feature Fusion; and (e) Fake News Detection. The text data is collected from the standard datasets. The collected text data is preprocessed and given into the feature extraction, which is done by using bidirectional encoder representations from transformers (BERT), transformer networks, and seq2seq network for extracting the three sets of language text features. These extracted feature sets are inserted into the weighted stacked feature fusion model, where the three sets of extracted features are integrated with the optimized weights that are acquired through the enhanced osprey optimization algorithm (EOOA). Finally, these resultant features are given to multi‐scale atrous convolution‐based one‐dimensional convolutional neural network with dilated long short‐term memory (MACNN‐DLSTM) for detecting the fake news. Throughout the result analysis, the experimentation is conducted based on the standard Tamil and Hindi datasets. Moreover, the developed model shows 92% for Hindi datasets and 96% for Tamil datasets which shows effective performance regarding accuracy measures. The experimental analysis is carried out by comparing with the conventional algorithms and detection techniques to showcase the efficiency of the developed regional language‐based fake news detection model.
{"title":"Adaptive weighted feature fusion for multiscale atrous convolution‐based 1DCNN with dilated LSTM‐aided fake news detection using regional language text information","authors":"V Rathinapriya, J. Kalaivani","doi":"10.1111/exsy.13665","DOIUrl":"https://doi.org/10.1111/exsy.13665","url":null,"abstract":"The people in the world rely on social media for gathering news, and it is mainly because of the development of technology. The approaches employed in natural language processing are still deficient in judgement factors, and these techniques frequently rely upon political or social circumstances. Numerous low‐level communities in the area are curious after experiencing the negative effects caused by the spread of false information in different sectors. Low‐resource languages are still distracted, because these techniques are extensively employed in the English language. This work aims to provide an analysis of regional language fake news and develop a referral system with advanced techniques to identify fake news in Hindi and Tamil. This proposed model includes (a) Regional Language Text Collection; (b) Text preprocessing; (c) Feature Extraction; (d) Weighted Stacked Feature Fusion; and (e) Fake News Detection. The text data is collected from the standard datasets. The collected text data is preprocessed and given into the feature extraction, which is done by using bidirectional encoder representations from transformers (BERT), transformer networks, and seq2seq network for extracting the three sets of language text features. These extracted feature sets are inserted into the weighted stacked feature fusion model, where the three sets of extracted features are integrated with the optimized weights that are acquired through the enhanced osprey optimization algorithm (EOOA). Finally, these resultant features are given to multi‐scale atrous convolution‐based one‐dimensional convolutional neural network with dilated long short‐term memory (MACNN‐DLSTM) for detecting the fake news. Throughout the result analysis, the experimentation is conducted based on the standard Tamil and Hindi datasets. Moreover, the developed model shows 92% for Hindi datasets and 96% for Tamil datasets which shows effective performance regarding accuracy measures. The experimental analysis is carried out by comparing with the conventional algorithms and detection techniques to showcase the efficiency of the developed regional language‐based fake news detection model.","PeriodicalId":51053,"journal":{"name":"Expert Systems","volume":null,"pages":null},"PeriodicalIF":3.3,"publicationDate":"2024-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141551442","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Eduardo M. Rodrigues, Yassine Baghoussi, João Mendes‐Moreira
Deep learning models are widely used in multivariate time series forecasting, yet, they have high computational costs. One way to reduce this cost is by reducing data dimensionality, which involves removing unimportant or low importance information with the proper method. This work presents a study on an explainability feature selection framework composed of four methods (IMV‐LSTM Tensor, LIME‐LSTM, Average SHAP‐LSTM, and Instance SHAP‐LSTM) aimed at using the LSTM black‐box model complexity to its favour, with the end goal of improving the error metrics and reducing the computational cost on a forecast task. To test the framework, three datasets with a total of 101 multivariate time series were used, with the explainability methods outperforming the baseline methods in most of the data, be it in error metrics or computation time for the LSTM model training.
{"title":"KDBI special issue: Explainability feature selection framework application for LSTM multivariate time‐series forecast self optimization","authors":"Eduardo M. Rodrigues, Yassine Baghoussi, João Mendes‐Moreira","doi":"10.1111/exsy.13674","DOIUrl":"https://doi.org/10.1111/exsy.13674","url":null,"abstract":"Deep learning models are widely used in multivariate time series forecasting, yet, they have high computational costs. One way to reduce this cost is by reducing data dimensionality, which involves removing unimportant or low importance information with the proper method. This work presents a study on an explainability feature selection framework composed of four methods (IMV‐LSTM Tensor, LIME‐LSTM, Average SHAP‐LSTM, and Instance SHAP‐LSTM) aimed at using the LSTM black‐box model complexity to its favour, with the end goal of improving the error metrics and reducing the computational cost on a forecast task. To test the framework, three datasets with a total of 101 multivariate time series were used, with the explainability methods outperforming the baseline methods in most of the data, be it in error metrics or computation time for the LSTM model training.","PeriodicalId":51053,"journal":{"name":"Expert Systems","volume":null,"pages":null},"PeriodicalIF":3.3,"publicationDate":"2024-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141553055","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Accurate prediction of gastric cancer survival state is one of great significant tasks for clinical decision‐making. Many advanced machine learning classification techniques have been applied to predict the survival status of cancer patients in three or 5 years, however, many of them have a low sensitivity because of class imbalance. This is a non‐negligible problem due to the poor prognosis of gastric cancer patients. Furthermore, models in the medical domain require strong interpretability to increase their applicability. Due to the better performance and interpretability of the XGBoost model, we design a loss function taking into account cost sensitive and focal loss from the algorithm level for XGBoost to deal with the imbalance problem. We apply the improved model into the prediction of the survival status of gastric cancer patients and analyse the important related features. We use two types of indicators to evaluate the model, and we also design the confusion matrix of two models' predictive results to compare two models. The results show that the improved model has better performance. Furthermore, we calculate the importance of features related to survival with three different time periods and analyse their evolution, which are consistent with existing clinical research or further expand their research conclusions. These all support for clinically relevant decision‐making and has the potential to expand into survival prediction of other cancer patients.
{"title":"Imbalanced survival prediction for gastric cancer patients based on improved XGBoost with cost sensitive and focal loss","authors":"Liangchen Xu, Chonghui Guo","doi":"10.1111/exsy.13666","DOIUrl":"https://doi.org/10.1111/exsy.13666","url":null,"abstract":"Accurate prediction of gastric cancer survival state is one of great significant tasks for clinical decision‐making. Many advanced machine learning classification techniques have been applied to predict the survival status of cancer patients in three or 5 years, however, many of them have a low sensitivity because of class imbalance. This is a non‐negligible problem due to the poor prognosis of gastric cancer patients. Furthermore, models in the medical domain require strong interpretability to increase their applicability. Due to the better performance and interpretability of the XGBoost model, we design a loss function taking into account cost sensitive and focal loss from the algorithm level for XGBoost to deal with the imbalance problem. We apply the improved model into the prediction of the survival status of gastric cancer patients and analyse the important related features. We use two types of indicators to evaluate the model, and we also design the confusion matrix of two models' predictive results to compare two models. The results show that the improved model has better performance. Furthermore, we calculate the importance of features related to survival with three different time periods and analyse their evolution, which are consistent with existing clinical research or further expand their research conclusions. These all support for clinically relevant decision‐making and has the potential to expand into survival prediction of other cancer patients.","PeriodicalId":51053,"journal":{"name":"Expert Systems","volume":null,"pages":null},"PeriodicalIF":3.3,"publicationDate":"2024-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141551444","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Daniel González Cortés, Enrique Onieva, Iker Pastor, Laura Trinchera, Jian Wu
While machine learning's role in financial trading has advanced considerably, algorithmic transparency and explainability challenges still exist. This research enriches prior studies focused on high‐frequency financial data prediction by introducing an explainable reinforcement learning model for portfolio management. This model transcends basic asset prediction, formulating concrete, actionable trading strategies. The methodology is applied in a custom trading environment mimicking the CAC‐40 index's financial conditions, allowing the model to adapt dynamically to market changes based on iterative learning from historical data. Empirical findings reveal that the model outperforms an equally weighted portfolio in out‐of‐sample tests. The study offers a dual contribution: it elevates algorithmic planning while significantly boosting transparency and interpretability in financial machine learning. This approach tackles the enduring ‘black‐box’ issue and provides a holistic, transparent framework for managing investment portfolios.
{"title":"Portfolio construction using explainable reinforcement learning","authors":"Daniel González Cortés, Enrique Onieva, Iker Pastor, Laura Trinchera, Jian Wu","doi":"10.1111/exsy.13667","DOIUrl":"https://doi.org/10.1111/exsy.13667","url":null,"abstract":"While machine learning's role in financial trading has advanced considerably, algorithmic transparency and explainability challenges still exist. This research enriches prior studies focused on high‐frequency financial data prediction by introducing an explainable reinforcement learning model for portfolio management. This model transcends basic asset prediction, formulating concrete, actionable trading strategies. The methodology is applied in a custom trading environment mimicking the CAC‐40 index's financial conditions, allowing the model to adapt dynamically to market changes based on iterative learning from historical data. Empirical findings reveal that the model outperforms an equally weighted portfolio in out‐of‐sample tests. The study offers a dual contribution: it elevates algorithmic planning while significantly boosting transparency and interpretability in financial machine learning. This approach tackles the enduring ‘black‐box’ issue and provides a holistic, transparent framework for managing investment portfolios.","PeriodicalId":51053,"journal":{"name":"Expert Systems","volume":null,"pages":null},"PeriodicalIF":3.3,"publicationDate":"2024-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141551443","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Object tracking is an essential component of computer vision and plays a significant role in various practical applications. Recently, transformer‐based trackers have become the predominant method for tracking due to their robustness and efficiency. However, existing transformer‐based trackers typically focus solely on the template features, neglecting the interactions between the search features and the template features during the tracking process. To address this issue, this article introduces a multi‐head cross‐attention transformer for visual tracking (MCTT), which effectively enhance the interaction between the template branch and the search branch, enabling the tracker to prioritize discriminative feature. Additionally, an auxiliary segmentation mask head has been designed to produce a pixel‐level feature representation, enhancing and tracking accuracy by predicting a set of binary masks. Comprehensive experiments have been performed on benchmark datasets, such as LaSOT, GOT‐10k, UAV123 and TrackingNet using various advanced methods, demonstrating that our approach achieves promising tracking performance. MCTT achieves an AO score of 72.8 on the GOT‐10k.
{"title":"An efficient object tracking based on multi‐head cross‐attention transformer","authors":"Jiahai Dai, Huimin Li, Shan Jiang, Hongwei Yang","doi":"10.1111/exsy.13650","DOIUrl":"https://doi.org/10.1111/exsy.13650","url":null,"abstract":"Object tracking is an essential component of computer vision and plays a significant role in various practical applications. Recently, transformer‐based trackers have become the predominant method for tracking due to their robustness and efficiency. However, existing transformer‐based trackers typically focus solely on the template features, neglecting the interactions between the search features and the template features during the tracking process. To address this issue, this article introduces a multi‐head cross‐attention transformer for visual tracking (MCTT), which effectively enhance the interaction between the template branch and the search branch, enabling the tracker to prioritize discriminative feature. Additionally, an auxiliary segmentation mask head has been designed to produce a pixel‐level feature representation, enhancing and tracking accuracy by predicting a set of binary masks. Comprehensive experiments have been performed on benchmark datasets, such as LaSOT, GOT‐10k, UAV123 and TrackingNet using various advanced methods, demonstrating that our approach achieves promising tracking performance. MCTT achieves an AO score of 72.8 on the GOT‐10k.","PeriodicalId":51053,"journal":{"name":"Expert Systems","volume":null,"pages":null},"PeriodicalIF":3.3,"publicationDate":"2024-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141510284","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}