Context and background: Complex fuzzy theory has a strong practical implication in many real-world applications. Complex Fuzzy Inference System (CFIS) is a powerful technique to overcome the challenges of uncertain, periodic data. However, a question is raised for CFIS: How can we deduce and predict the result in case there is little knowledge about data information and rule base? This is significance because many real applications do not have enough knowledge of rule base for inference so that the performance of systems may be low. Thus, it is necessary to have an approximate reasoning method to represent and derive final results. Motivation: Recently, the Mamdani Complex Fuzzy Inference System (M-CFIS) has been proposed with a specific inference mechanism according to the Mamdani type. A new improvement so-called the Mamdani Complex Fuzzy Inference System with Rule Reduction (M-CFIS-R) has been designed to utilize granular computing with complex similarity measures to reduce the rule base so as to gain better performance in decision-making problems. However in M-CFIS-R, testing data are checked by matching with each rule in the rule base, which leads to a high cost of computational time. Besides, if the testing data contain records that are not inferred by the rule base, the output cannot be generated. This happens in real commerce systems in which the rule base is small at the time of creation and needs to feed with new rules. Methodology: In order to handle those issues, this paper first time proposes the Fuzzy Knowledge Graph to represent the rule base in terms of linguistic labels and their relationships according to the rule set. An adjacent matrix of Fuzzy Knowledge Graph is generated for inference. When a record in the Testing dataset is given, it would be fuzzified and labelled. Each component in the record is checked with the Fuzzy Knowledge Graph by the inference mechanism in approximate reasoning called Fast Inference Search Algorithm. Then, we derive the label of the new record by the Max-Min operator. Besides, we also propose four extensions of Mamdani Complex Fuzzy Inference System Rule Reduction including Sugeno Complex Fuzzy Inference Systems, Tsukamoto Complex Fuzzy Inference Systems, Complex Fuzzy Measures and Complex Fuzzy Integrals in M-CFIS-R. Results: The experiments on the UCI Machine Learning datasets show that the proposed method classifies samples as correctly as M-CFIS-R with very lower run time (6.45 times on average). The experiments are performed through many tests via 2 main scenarios. Conclusion: The proposed system has good performance in reducing the computational time of inference with acceptable accuracy. It has ability to work with systems having limited knowledge and rule base.

ChatGPT plays significant roles in the third decade of the 21st Century. Smart cities applications can be integrated with ChatGPT in various fields. This research proposes an approach for developing large language models using generative artificial intelligence models suitable for small- and medium-sized enterprises with limited hardware resources. There are many generative AI systems in operation and in development. However, the technological, human, and financial resources required to develop generative AI systems are impractical for small- and medium-sized enterprises. In this study, we present a proposed approach to reduce training time and computational cost that is designed to automate question–response interactions for specific domains in smart cities. The proposed model utilises the BLOOM approach as its backbone for using generative AI to maximum the effectiveness of small- and medium-sized enterprises. We have conducted a set of experiments on several datasets associated with specific domains to validate the effectiveness of the proposed model. Experiments using datasets for the English and Vietnamese languages have been combined with model training using low-rank adaptation to reduce training time and computational cost. In comparative experimental testing, the proposed model outperformed the ‘Phoenix’ multilingual chatbot model by achieving a 92% performance compared to ‘ChatGPT’ for the English benchmark.

Precipitation nowcasting is one of the main tasks of weather forecasting that aims to predict rainfall events accurately, even in low-rainfall regions. It has been observed that few studies have been devoted to predicting future radar echo images in a reasonable time using the deep learning approach. In this paper, we propose a novel approach, RainPredRNN, which is the combination of the UNet segmentation model and the PredRNN_v2 deep learning model for precipitation nowcasting with weather radar echo images. By leveraging the abilities of the contracting-expansive path of the UNet model, the number of calculated operations of the RainPredRNN model is significantly reduced. This result consequently offers the benefit of reducing the processing time of the overall model while maintaining reasonable errors in the predicted images. In order to validate the proposed model, we performed experiments on real reflectivity fields collected from the Phadin weather radar station, located at Dien Bien province in Vietnam. Some credible quality metrics, such as the mean absolute error (MAE), the structural similarity index measure (SSIM), and the critical success index (CSI), were used for analyzing the performance of the model. It has been certified that the proposed model has produced improved performance, about 0.43, 0.95, and 0.94 of MAE, SSIM, and CSI, respectively, with only 30% of training time compared to the other methods.

The next generation of E-Government and healthcare has the potential to increase the more intelligent governance with improvements in transparency, accountability, efficiency, and effectiveness. It enables organizations to use the benefits of information via big data analysis to settle the difficulties effectively. Big Data has emerged which plays a significant role in many sectors around the world. Global trends in taking advantage of the benefits from big data are considered with an overview of the US, European Union, and several developing countries. To deeply understand the utilization of big data in several domains, this study has presented a brief survey of key concepts (such as IoT-enabled data, blockchain-enabled data, and intelligent systems data) to deeply understand the utilization of big data in several domains. Our analysis sets out also the similarities and differences in these concepts. We have also surveyed state-of-the-art technologies including cloud computing, multi-cloud, webservice, and microservice which are used to exploit potential benefits of big data analytics. Furthermore, some typical big data frameworks are surveyed and a big data framework for E-Government is also proposed. Open research questions and challenges are highlighted (for researchers and developers) following our review. Our goal in presenting the novel concepts presented in this article is to promote creative ideas in the research endeavor to perform efficaciously next-generation E-Government in the context of Industry 4.0.

Deep learning methods predicated on convolutional neural networks and graph neural networks have enabled significant improvement in node classification and prediction when applied to graph representation with learning node embedding to effectively represent the hierarchical properties of graphs. An interesting approach (DiffPool) utilises a differentiable graph pooling technique which learns ‘differentiable soft cluster assignment’ for nodes at each layer of a deep graph neural network with nodes mapped on sets of clusters. However, effective control of the learning process is difficult given the inherent complexity in an ‘end-to-end’ model with the potential for a large number parameters (including the potential for redundant parameters). In this paper, we propose an approach termed FPool, which is a development of the basic method adopted in DiffPool (where pooling is applied directly to node representations). Techniques designed to enhance data classification have been created and evaluated using a number of popular and publicly available sensor datasets. Experimental results for FPool demonstrate improved classification and prediction performance when compared to alternative methods considered. Moreover, FPool shows a significant reduction in the training time over the basic DiffPool framework.