Journal of Intelligent & Fuzzy Systems - Volume 46, issue 3

Authors: Ouyang, Zhiyuan | Wan, Yanling | Zhang, Tao | Wu, Wen-Ze

Article Type: Research Article

Abstract: The introduction of fractional order accumulation has played a crucial role in the development of grey forecasting methods. However, accurately identifying a single fractional order accumulation for modeling diverse sequences is challenging due to the dependence of different fractional order accumulations on data structure over time. To address this issue, we propose a novel fractional grey model abbreviated as FGMMA, incorporating a model averaging method. The new model combines existing fractional grey models by using four judgment criteria, including Akaike information criteria, Bayesian information criteria, Mallows criteria, and Jackknife criteria. Meanwhile, the cutting-edge algorithm named breed particle swarm optimization is …employed to search the optimal fractional order for each candidate model to enhance the effectiveness of the designed model. Subsequently, we conduct a Monte Carlo simulation for verification and validation purposes. Finally, empirical analysis based on energy consumption in three countries is conducted to verify the applicability of the proposed model. Compared with other benchmark models, we can conclude that the proposed model outperforms the other competitive models. Show more

Keywords: Grey forecasting model, fractional order accumulation, model averaging, breed particle swarm optimization

DOI: 10.3233/JIFS-237479

Citation: Journal of Intelligent & Fuzzy Systems, vol. 46, no. 3, pp. 6479-6490, 2024

Authors: Zhang, Yihao | Wang, Yuhao | Lan, Pengxiang | Xiang, Haoran | Zhu, Junlin | Yuan, Meng

Article Type: Research Article

Abstract: Conversational recommender systems use natural language conversations to elicit user preferences and recommend items proactively. Existing methods based on graph neural networks have been proven to be effective in exploiting knowledge graphs. However, node positions are often treated as constants, which leads to the neglect of graph connectivity due to fuzzy processing. In addition, although the transformer has significant advantages in understanding the text, its secondary computational complexity may be incapable when dealing with long texts. In order to solve these problems, we propose an additive positional conversational recommender model called APCR. This model converts the pair product of transformer …into a linear operation, and uses the Laplacian eigenvector to build a location graph. The extended graph neural network captures the topology structure of the location knowledge graph. Specifically, we design an encoder based on additive attention to break through the bottleneck of long text. Furthermore, we develop a recommendation model based on a positional graph neural network to match items with dialogue context, thereby capturing the graph topology. Extensive experiments on the REDIAL dataset show significant improvements in our proposed model over the state-of-the-art methods in recommendation and dialogue generation evaluations. Show more

Keywords: Interactive recommender systems, graph neural networks, knowledge graphs, additive attention

DOI: 10.3233/JIFS-230905

Citation: Journal of Intelligent & Fuzzy Systems, vol. 46, no. 3, pp. 6491-6503, 2024

Authors: Qian, Jin | Wang, Taotao | Lu, Yuehua | Yu, Ying

Article Type: Research Article

Abstract: Multi-granularity hesitant fuzzy linguistic terms set is an effective expression of linguistic information, which can utilize some fuzzy linguistic terms to evaluate various common qualitative information and plays an important role when experts provide linguistic information to express hesitancy. Since the alternative description in the decision-making information system is characterized by multi-granularity, uncertainty, and vagueness, this paper proposes a multi-granularity hesitant fuzzy linguistic decision-making VIKOR method based on entropy weight and information transformation. Specifically, this paper firstly adopts fuzzy information entropy to obtain the weights of different attributes and introduces a multi-granularity hesitant fuzzy linguistic term set conversion method to …realize the semantic information conversion between different granularities. Then for the converted affiliation linguistic decision matrix, the entropy weighting method is used to obtain the weights of different affiliation granularity layers, and a weight optimization VIKOR method based on the affiliation linguistic decision matrix is further proposed to rank the alternatives. Finally, the feasibility of the proposed method verified by arithmetic examples, experimental analysis is carried out in terms of parameter sensitivity analysis and comparison with other methods. The experimental results prove the rationality and effectiveness of the proposed method. Show more

Keywords: Multi-granularity hesitant fuzzy term set, affiliation degree, information transformation, VIKOR method

DOI: 10.3233/JIFS-237951

Citation: Journal of Intelligent & Fuzzy Systems, vol. 46, no. 3, pp. 6505-6516, 2024

Authors: Prasath, J.S. | Shyja, V. Irine | Chandrakanth, P. | Kumar, Boddepalli Kiran | Raja Basha, Adam

Article Type: Research Article

Abstract: Now, the Cyber security is facing unprecedented difficulties as a result of the proliferation of smart devices in the Internet of Things (IoT) environment. The rapid growth in the number of Internet users over the past two decades has increased the need for cyber security. Users have provided new opportunities for attackers to do harm. Limited security budgets leave IoT devices vulnerable and easily hacked to launch distributed denial-of-service (DDoS) attacks, with disastrous results. Unfortunately, due to the unique nature of the Internet of Things environment, most security solutions and intrusion detection systems (IDS) cannot be directly adapted to the …IoT with acceptable security performance and are vulnerable to various attacks that do not benefit. In this paper we propose an optimal secure defense mechanism for DDoS in IoT network using feature optimization and intrusion detection system (OSD-IDS). In OSD-IDS mechanism, first we introduce an enhanced ResNet architecture for feature extraction which extracts more deep features from given traffic traces. An improved quantum query optimization (IQQO) algorithm for is used feature selection to selects optimal best among multiple features which reduces the data dimensionality issues. The selected features have given to the detection and classification module to classify the traffic traces are affected by intrusion or not. For this, we design a fast and accurate intrusion detection mechanism, named as hybrid deep learning technique which combines convolutional neural network (CNN) and diagonal XG boosting (CNN-DigXG) for the fast and accurate intrusion detection in IoT network. Finally, we validate the performance of proposed technique by using different benchmark datasets are BoNeSi-SlowHTTPtest and CIC-DDoS2019. The simulation results of proposed IDS mechanism are compared with the existing state-of-art IDS mechanism and analyze the performance with respects to different statistical measures. The results show that the DDoS detection accuracy of proposed OSD-IDS mechanism is high as 99.476% and 99.078% for BoNeSi-SlowHTTPtest, CICDDoS2019, respectively. Show more

Keywords: Defense mechanism, DDoS intrusion, intrusion detection system, feature selection, IoT

DOI: 10.3233/JIFS-235529

Citation: Journal of Intelligent & Fuzzy Systems, vol. 46, no. 3, pp. 6517-6534, 2024

Authors: Ye, Qing | Song, Zihan | Zhao, Yuqi | Zhang, Yongmei

Article Type: Research Article

Abstract: Video anomaly detection refers to the automatic identification of abnormal behaviors, objects, or events in videos. However, current methods for anomaly detection based on original frames lack a comprehensive understanding of the importance of foreground information, making it challenging to efficiently address video anomaly detection in the presence of complex background interference. In this paper, we propose a video anomaly detection algorithm based on Background Separation Network (BSN) to address this issue. Firstly, we utilize a video stabilization algorithm to reduce video jitter and enhance the quality of input video frames. Secondly, BSN shifts the focus from the entire frame …to the foreground region with higher anomaly detection value. BSN utilizes the motion pixel distribution of the video as the basis for foreground extraction, enabling pixel-level background separation to obtain more accurate and complete foreground targets. Lastly, a certain proportion of foreground targets in the foreground image are masked as background, reducing the interference caused by redundant targets on the detection results. The proposed method achieves an accuracy of 96.2% on the UCSD ped2 dataset, demonstrating its effectiveness. This method contributes to accurately detecting abnormal behaviors in real-world surveillance videos to protect the safety of public lives and assets. Show more

Keywords: Video anomaly detection, auto encoder, background separation network, video jitter elimination

DOI: 10.3233/JIFS-235717

Citation: Journal of Intelligent & Fuzzy Systems, vol. 46, no. 3, pp. 6535-6551, 2024

Authors: Sheikh, Ansar Isak | Sadish Sendil, M. | Sridhar, P. | Thariq Hussan, M.I. | Abidin, Shafiqul | Kumar, Ravi | Irshad, Reyazur Rashid | Muniyandy, Elangovan | Phani Kumar, Solleti

Article Type: Research Article

Abstract: This article has been retracted. A retraction notice can be found at https://doi.org/10.3233/JIFS-219433 .

DOI: 10.3233/JIFS-237474

Citation: Journal of Intelligent & Fuzzy Systems, vol. 46, no. 3, pp. 6553-6564, 2024

Authors: Runkler, Thomas A.

Article Type: Research Article

Abstract: Pairwise fuzzy preference matrices can be constructed using expert ratings. The number of pairwise preference values to be specified by the experts increases quadratically with the number of options. Consistency (transitivity) allows to reduce this quadratic complexity to linear complexity which makes this approach feasible also for large scale applications. Preference values are usually expected to be on a fixed finite interval. Additive preference is defined on such a finite interval. However, completing preference matrices using additive consistency may yield preferences outside this finite interval. Multiplicative preference is defined on an infinite interval and is therefore not suitable here. …To overcome this problem we extend the concept of consistency beyond additive and multiplicative to arbitrary commutative, associative, and invertible operators. Infinitely many of such operators induce infinitely many types of consistency. As one example, we examine Einstein consistency, which is induced by the Einstein sum operator. Completing preference matrices using Einstein consistency always yields preferences inside the finite interval, which yields the first method that allows to construct large scale finite preference matrices using expert ratings. A case study with the real–world car preference data set indicates that Einstein consistency also yields more accurate preference estimates than additive or multiplicative consistency. Show more

Keywords: Fuzzy preference relations, consistent preference, additive preference, multiplicative preference, Einstein sum

DOI: 10.3233/JIFS-224179

Citation: Journal of Intelligent & Fuzzy Systems, vol. 46, no. 3, pp. 6565-6576, 2024

Authors: Bu, Yanbin | Chen, Ting | Duan, Hongxiu | Liu, Mei | Xue, Yandan

Article Type: Research Article

Abstract: In the modern world, structured and semi-structured knowledge bases hold a considerable amount of data. There-fore, people who are familiar with formal query languages should not be the only ones who can efficiently and clearly query them. Semantic Parsing (SP) is converting natural language utterances into formal meaning representations. The paper suggests a model for SP that uses a novel method of utilizing the Semi-Supervised Generative Adversarial Network (SS-GAN) to enhance the classifier performance. The proposed SS-GAN extends the fine-tuning of word embedding architectures using unlabeled examples in a generative adversarial environment. We provide a regularization strategy for addressing the …mode missing problem and unstable training in SS-GAN. The main viewpoint is to use the extracted feature vectors from the discriminator. Hence, the generator produces outputs by aiding the discriminator’s learned features. A reconstruction loss is added to the loss function of the SS-GAN to drive the genera-tor to reconstruct outputs from the discriminator’s features, hence steering the generator toward actual data configurations. The proposed reconstruction loss improves the performance of SS-GAN, produces high-quality outputs, and may be combined with other regularization loss functions to improve the performance of diverse GANs. We employ BERT word embedding for our model, which can be included in a downstream task and fine-tuned as a model, while the pre-trained BERT model can capture various linguistic properties. We examine the suggested model using the WikiSQL and SparC datasets, and the analysis findings reveal our model outperforms its rivals. The findings from our experiments indicate that the need for labeled samples can be minimized, down to as few as 100 instances, while still achieving commendable classification outcomes. Show more

Keywords: Semantic parsing, generative adversarial network, semi-supervised learning, BERT

DOI: 10.3233/JIFS-233212

Citation: Journal of Intelligent & Fuzzy Systems, vol. 46, no. 3, pp. 6577-6588, 2024

Authors: Li, Feng | Zhu, Mozhong | Lin, Ling

Article Type: Research Article

Abstract: This article has been retracted. A retraction notice can be found at https://doi.org/10.3233/JIFS-219433 .

DOI: 10.3233/JIFS-234686

Citation: Journal of Intelligent & Fuzzy Systems, vol. 46, no. 3, pp. 6589-6605, 2024

Authors: Ma, Fanglan | Zhu, Changsheng | Liu, Dukui

Article Type: Research Article

Abstract: Knowledge tracing (KT), which aims to trace human knowledge learning process by using machines, has widely applied in online learning systems. It dynamically models student’s knowledge states in relation to different learning factors through their learning interactions. Recently, KT has attracted many researches attention due to its good performance to using deep learning. Although most of KT models have shown outstanding results, they have limitations: either ignore the human cognitive law and learning behavior, or lack the ability to go deeper modeling to trace knowledge state. In this paper, we propose a deeper knowledge tracking model integrating cognitive theory and …learning behavior (CLDKT). It united the advantages of memory network and recurrent neural network of the existing deep learning KT models for modeling student learning. To better implement CLDKT, we add the residual network (ResNet) to realize the deep modeling of learning behaviors. Extensive experiments on three open benchmark datasets to evaluate our model. Experimental results demonstrate that (I) CLDKT outperforms the state-of-the-art KT models on students’ performance prediction. (II) CLDKT can deeper modeling to trace knowledge state owing to the ResNet import. (III) CLDKT has better interpretability and predictability, which proves the effectiveness of the knowledge tracing model integrating cognitive law and learning behavior. Show more

Keywords: Knowledge tracing, cognitive law, learning behavior, ResNet, deep learning

DOI: 10.3233/JIFS-235723

Citation: Journal of Intelligent & Fuzzy Systems, vol. 46, no. 3, pp. 6607-6617, 2024