Journal of Intelligent & Fuzzy Systems - Volume Pre-press, issue Pre-press

Authors: Zhao, Liang | Wang, Jiawei | Liu, Shipeng | Yang, Xiaoyan

Article Type: Research Article

Abstract: Tunnels water leakage detection in complex environments is difficult to detect the edge information due to the structural similarity between the region of water seepage and wet stains. In order to address the issue, this study proposes a model comprising a multilevel transformer encoder and an adaptive multitask decoder. The multilevel transformer encoder is a layered transformer to extract the multilevel characteristics of water leakage information, and the adaptive multitask decoder comprises the adaptive network branches. The adaptive network branches generate the ground truths of wet stains and water seepage through the threshold value and transmit them to the network …for training. The converged network, the U-net, fuses coarse images from the adaptive multitask decoder, and the fusion images are the final segmentation results of water leakage in tunnels. The experimental results indicate that the proposed model achieves 95.1% Dice and 90.4% MIOU, respectively. This proposed model demonstrates a superior level of precision and generalization when compared to other related models. Show more

Keywords: Water leakage, multilevel transformer encoder, adaptive multitask decoder, adaptive network branches, converged network

DOI: 10.3233/JIFS-224315

Citation: Journal of Intelligent & Fuzzy Systems, vol. Pre-press, no. Pre-press, pp. 1-15, 2023

Authors: Luo, Binghui | Liu, Xin | Qin, Long | Jiao, Xiaolong | Li, Wengui

Article Type: Research Article

Abstract: The short text matching models can be roughly divided into representation-based and interaction-based approaches. However, current representation-based text matching models often lack the ability to handle sentence pairs and typically only perform feature interactions at the network’s top layer, which can lead to a loss of semantic focus. The interactive text matching model has significant shortcomings in extracting differential information between sentences and may ignore global information. To address these issues, this article proposes a model structure that combines a dual-tower architecture with an interactive component, which compensates for their respective weaknesses in extracting sentence semantic information. Simultaneously, a method …for integrating semantic information is proposed, enabling the model to capture both the interactive information between sentence pairs and the differential information between sentences, thereby addressing the issues with the aforementioned approaches. In the process of network training, a combination of cross-entropy and cosine similarity is used to calculate the model loss. The model is optimized to a stable state. Experiments on the commonly used datasets of QQP and MRPC validate the effectiveness of the proposed model, and its performance is stably improved. Show more

Keywords: Short text matching, representational structure, interactive structure, BERT, multi-angle information

DOI: 10.3233/JIFS-230268

Citation: Journal of Intelligent & Fuzzy Systems, vol. Pre-press, no. Pre-press, pp. 1-13, 2024

Authors: Diao, Xiu-Li | Zhang, Hao-Ran | Zeng, Qing-Tian | Song, Zheng-Guo | Zhao, Hua

Article Type: Research Article

Abstract: At present, the Chinese text field is facing challenges from low resource issues such as data scarcity and annotation difficulties. Moreover, in the domain of cigarette tasting, cigarette tasting texts tend to be colloquial, making it difficult to obtain valuable and high-quality tasting texts. Therefore, in this paper, we construct a cigarette tasting dataset (CT2023) and propose a novel Chinese text classification method based on ERNIE and Comparative Learning for Low-Resource scenarios (ECLLR). Firstly, to address the issues of limited vocabulary diversity and sparse features in cigarette tasting text, we utilize Term Frequency-Inverse Document Frequency (TF-IDF) to extract key terms, …supplementing the discriminative features of the original text. Secondly, ERNIE is employed to obtain sentence-level vector embedding of the text. Finally, contrastive learning model is used to further refine the text after fusing the keyword features, thereby enhancing the performance of the proposed text classification model. Experiments on the CT2023 dataset demonstrate an accuracy rate of 96.33% for the proposed method, surpassing the baseline model by at least 11 percentage points, and showing good text classification performance. It is thus clear that the proposed approach can effectively provide recommendations and decision support for cigarette production processes in tobacco companies. Show more

Keywords: Low-resource, Cigarette Tasting, Contrastive Learning, Text classification

DOI: 10.3233/JIFS-237816

Citation: Journal of Intelligent & Fuzzy Systems, vol. Pre-press, no. Pre-press, pp. 1-15, 2024

Authors: Ledesma Roque, Diana Anahí | Kolesnikova, Olga | Menchaca Méndez, Ricardo

Article Type: Research Article

Abstract: This study addresses the issue of semantic similarity in sentences using the BERT model through various aggregation techniques, such as max-pooling, mean-pooling, and an LSTM network applied to the output of the BERT model. Subsequently, the linguistic interpretability of the BERT-Base transformer model is analyzed through the unsupervised learning approach, specifically through dimensionality reduction using autoencoders and clustering algorithms, utilizing the representation of the classification token CLS. The results highlight that the CLS classification token achieves better abstractions than the proposed methods. In terms of interpretability, it is observed that sequence length is relevant in the early layers, with …a gradual decrease across the layers. Additionally, attention to semantic similarity is concentrated in the intermediate and upper layers, especially in layers 6, 8, 9, and 10. All these findings were obtained by addressing the semantic similarity task using the STS-Benchmark dataset. Show more

Keywords: Linguistic interpretability, aggregation methods, unsupervised learning, attention mechanisms, token CLS

DOI: 10.3233/JIFS-219359

Citation: Journal of Intelligent & Fuzzy Systems, vol. Pre-press, no. Pre-press, pp. 1-14, 2024

Authors: Cardoso-Moreno, Marco A. | Luján-García, Juan Eduardo | Yáñez-Márquez, Cornelio

Article Type: Research Article

Abstract: In this study, a thorough analysis of the proposed approach in the context of emotion classification using both single-modal (A-13sbj) and multi-modal (B-12sbj) sets from the YAAD dataset was conducted. This dataset encompassed 25 subjects exposed to audiovisual stimuli designed to induce seven distinct emotional states. Electrocardiogram (ECG) and galvanic skin response (GSR) biosignals were collected and classified using two deep learning models, BEC-1D and ELINA, along with two different preprocessing techniques, a classical fourier-based filtering and an Empirical Mode Decomposition (EMD) approach. For the single-modal set, this proposal achieved an accuracy of 84.43±30.03, precision of 85.16±28.91, and F1-score of …84.06±29.97. Moreover, in the extended configuration the model maintained strong performance, yielding scores of 80.95±22.55, 82.44±24.34, and 79.91±24.55, respectively. Notably, for the multi-modal set (B-12sbj), the best results were obtained with EMD preprocessing and the ELINA model. This proposal achieved an improved accuracy, precision, and F1-score scores of 98.02±3.78, 98.31±3.31, and 97.98±3.83, respectively, demonstrating the effectiveness of this approach in discerning emotional states from biosignals. Show more

Keywords: Emotion classification, signal preprocessing, convolutional neural network, ECG, GSR, EMD

DOI: 10.3233/JIFS-219334

Citation: Journal of Intelligent & Fuzzy Systems, vol. Pre-press, no. Pre-press, pp. 1-9, 2024

Authors: Yigezu, Mesay Gemeda | Kolesnikova, Olga | Gelbukh, Alexander | Sidorov, Grigori

Article Type: Research Article

Abstract: The rise of social media and micro-blogging platforms has led to concerns about hate speech, its potential to incite violence, psychological trauma, extremist beliefs, and self-harm. We have proposed a novel model, Odio-BERT for detecting hate speech using a pretrained BERT language model. This specialized model is specifically designed for detecting hate speech in the Spanish language, and when compared to existing models, it consistently outperforms them. The study provides valuable insights into addressing hate speech in the Spanish language and explores the impact of domain tasks.

Keywords: BERT, hate speech, domain task, fine tune, Odio-BERT, Spanish

DOI: 10.3233/JIFS-219349

Citation: Journal of Intelligent & Fuzzy Systems, vol. Pre-press, no. Pre-press, pp. 1-12, 2024

Authors: Liang, Weijing | Xue, Ye | Xu, Jing

Article Type: Research Article

Abstract: With the increasing global disaster risks, constructing more inclusive, flexible, and resilient communities has become crucial for effectively carrying out disaster prevention, mitigation, and relief work. However, existing research on community resilience mostly focuses on the selection of key factors and the assessment of community resilience, lacking in-depth exploration of the interactions between factors and simulation studies of key paths. Therefore, this paper applies the Fuzzy Decision-Making Trial and Evaluation Laboratory (Fuzzy DEMATEL) method to select important factors of community resilience. Based on this, the maximum average difference entropy method is used to analyze the relationships and influence mechanisms among …different factors, thus identifying the key factors and key paths affecting community resilience. The Fuzzy Cognitive Map (FCM) is then used to simulate the paths. The study finds that factors of community resilience can be categorized as input, intermediary, and output types, and further analysis of their influence mechanisms reveals four key paths and four key factors. Through pathway simulation, different improvement states of community resilience are observed when triggering the input-type factors of the key paths. Therefore, under limited resources, a phased and systematic approach to enhancing community resilience should be adopted. The contribution of this study lies in providing a comprehensive analysis of factors and pathway selection methods, and through pathway simulation, it offers a scientific basis and decision support for improving and constructing community resilience in practice. Show more

Keywords: Fuzzy cognitive map, fuzzy DEMATEL, maximum average difference entropy method, community resilience, simulation analysis

DOI: 10.3233/JIFS-232234

Citation: Journal of Intelligent & Fuzzy Systems, vol. Pre-press, no. Pre-press, pp. 1-23, 2024

Authors: Zhang, Shuguang | Xie, Chengyuan | Zhang, Heng | Gong, Wenzheng | Liu, Lingjie | Zhi, Xuntao

Article Type: Research Article

Abstract: Graph Convolutional Networks (GCN) are prevalent techniques in collaborative filtering recommendations. However, current GCN-based approaches for collaborative filtering recommendation have limitations in effectively embedding neighboring nodes during node and neighbor information aggregation. Furthermore, weight allocation for the user (or item) representations after convolution of each layer is too uniform. To resolve these limitations, we propose a new Graph Convolutional Collaborative Filtering recommendation method based on temporal information during the node aggregation process (TA-GCCF). The method aggregates and propagates information using Gated Recurrent Units, while dynamically updating features based on the timing and sequence of interactions between nodes and their neighbors. …Concurrently, we have developed a convolution attention coefficient to ascertain the significance of embedding at distinct layers. Experiments on three benchmark datasets show that our method significantly outperforms the comparison methods in the accuracy of prediction. Show more

Keywords: Graph convolutional neural network, collaborative filtering, recommendation, gated recurrent units, temporal information

DOI: 10.3233/JIFS-238307

Citation: Journal of Intelligent & Fuzzy Systems, vol. Pre-press, no. Pre-press, pp. 1-10, 2024

Authors: Vela-Rincón, Virna V. | Mújica-Vargas, Dante | Luna-Álvarez, Antonio | Arenas Muñiz, Andrés Antonio | Cruz-Prospero, Luis A.

Article Type: Research Article

Abstract: Image segmentation is a very studied area, looking for the best clustering of pixels. However, it is sometimes a complicated task, especially when these pixels are at the edges of regions, where there is a gradient and it is difficult to decide to which region to assign it. Hesitating fuzzy sets (HFS) better describe these situations, allowing to have multiple possible values for each element, giving more flexibility. This type of sets has been mainly applied in decision-making problems, obtaining better results than other types of fuzzy sets. This research proposes a fast and automatic method based on fuzzy hesitant …clustering (FAHFC), which does not require parameters since it is capable of determining the number of clusters, using the Calinski-Harabasz index, in which the segmentation is performed, solving the initialization problem in clustering; it also proposes an alternative to construct the HFS through the use of fuzzy relations. The experiments show superiority in terms of clustering quality and convergence over some selected state-of-the-art algorithms. Show more

Keywords: Fuzzy clustering, hesitant fuzzy sets, image segmentation, Calinski-Harabasz index

DOI: 10.3233/JIFS-219370

Citation: Journal of Intelligent & Fuzzy Systems, vol. Pre-press, no. Pre-press, pp. 1-13, 2024

Authors: Milovanović, Vladimir | Aleksić, Aleksandar | Milenkov, Marjan | Sokolović, Vlada

Article Type: Research Article

Abstract: The paper aims to present a hybrid model for measuring the performance of business processes in complex organizations based on the subjective decision-making of expert teams. The subject of the research is finding ways to measure, analyze and improve the key performance indicators (KPIs) process. Obtaining the values of KPIs, which reflect the real state of the process, creates a basis for their ranking, i.e. insight into KPIs that are extremely important for the process as well as KPIs that are of lesser importance, but as such are not excluded from consideration because they are necessary for the beginning, realization …and completion of the process. The model was compiled through five phases and was tested through a case study in a real business organization, which deals with the maintenance of complex combat systems. The obtained results helped the management to take certain measures in order to improve the performance of the maintenance process. In the model, it is proposed to form two expert teams, which make assessments based on experience and express them in linguistic terms according to a predefined scale. Modeling of linguistic expressions is realized using intuitive fuzzy sets of a higher order, more precisely Fermatean fuzzy sets (FFS). Selecting KPIs, decomposing the process into sub-processes and assessing the relative importance of sub-processes is carried out by one team of experts, while another team carries out the assessment of KPIs at the level of each sub-process. Determining the relative importance of sub-processes is realized using the Delphi method extended to FFS while reaching a consensus. The measurement of process performance, i.e. the value of KPIs, is realized using Multi-Criteria Group Decision-Making (MCGDM), such as the ELECTRE method extended with FFS. The sensitivity analysis of the developed model is realized by uncertainty modeling with q-rung orthopair fuzzy sets. Show more

Keywords: Fermatean fuzzy set (FFSs), ELECTRE method, Delphi method, maintenance, performance

DOI: 10.3233/JIFS-238907

Citation: Journal of Intelligent & Fuzzy Systems, vol. Pre-press, no. Pre-press, pp. 1-17, 2024