International Journal of Uncertainty Fuzziness and Knowledge-Based Systems最新文献

Extracting the relations of two entities on the sentence-level has drawn increasing attention in recent years but remains facing great challenges on document-level, due to the inherent difficulty in recognizing the relations of two entities across multiple sentences. Previous works show that employing the graph convolutional neural network can help the model capture unstructured dependent information of entities. However, they usually employed the non-adaptive weight edges to build the correlation weight matrix which suffered from the problem of information redundancy and gradient disappearance. To solve this problem, we propose a deep gated graph reasoning model for document-level relation extraction, namely, BERT-GGNNs, which employ an improved gated graph neural network with a learnable correlation weight matrix to establish multiple deep gated graph reason layers. The proposed deep gated graph reasoning layers make the model easier to reasoning the relations between entities hidden in the document. Experiments show that the proposed model outperforms most of strong baseline models, and our proposed model is 0.3% and 0.3% higher than the famous LSR-BERT model on the F1 and Ing F1, respectively.

Better performance at a country level will provide benefits to the whole population. This issue has been studied from various perspectives using empirical methods. However, little effort has as yet been made to address the issue of endogeneity in the interrelationships between productive performance and its determinants. We address this issue by proposing a Two-Dimensional Fuzzy-Monte Carlo Analysis (2DFMC) approach. The joint use of stochastic and fuzzy approaches – within the ambit of 2DFMCA – offers methodological tools to mitigate epistemic uncertainty while increasing research validity and reproducibility: (i) preliminary performance assessment by fuzzy ideal solutions; and (ii) robust stochastic regression of the performance scores into the epistemic sources of uncertainty related to the levels of physical and human capitals measured in distinct countries at different epochs. By applying the proposed method to a sample of 23 countries for 1890–2018, our results show that the best and worst-performing countries were Norway and Portugal, respectively. We further found that the intensity of human capital and the age of equipment (capital stock) have different impacts on productive performance – it has been established that capital intensity and total factor productivity are influenced by productivity performance, which, in turn, has a negative impact on labor productivity and GDP per capita. Our analysis provides insights to enable government policies to coordinate productive performance and other macroeconomic indicators.

Algebraic operations on generalized hesitant fuzzy numbers are key tools to address the problems with decision uncertainty. In this paper, by studying the arithmetic operations on generalized trapezoidal hesitant fuzzy numbers, modified arithmetic operations are introduced for this class of numbers so that, using these arithmetic operations, the multiplication and division of two generalized trapezoidal hesitant fuzzy numbers are always generalized trapezoidal hesitant fuzzy numbers. Furthermore, a generalized trapezoidal hesitant fuzzy number raised to the power of a real number is a generalized trapezoidal hesitant fuzzy number, and in the defined division, the case where the denominator becomes zero is not considered. Numerical examples are used to show the shortcomings of the previous arithmetic operations as well as the efficiencies of the arithmetic operations proposed in this research for generalized trapezoidal hesitant fuzzy numbers. Finally, the application of the proposed new arithmetic operations to generalized trapezoidal hesitant fuzzy numbers in solving the generalized trapezoidal hesitant fully fuzzy equation is discussed.

Currently, social media networks such as Facebook and Twitter have evolved into valuable platforms for global communication. However, due to their extensive user bases, Twitter is often misused by illegitimate users engaging in illicit activities. While there are numerous research papers available that delve into combating illegitimate users on Twitter, a common shortcoming in most of these works is the failure to address the issue of class imbalance, which significantly impacts the effectiveness of spam detection. Few other research works that have addressed class imbalance have not yet applied bio-inspired algorithms to balance the dataset. Therefore, we introduce PSOB-U, a particle swarm optimization-based undersampling technique designed to balance the Twitter dataset. In PSOB-U, various classifiers and metrics are employed to select majority samples and rank them. Furthermore, an ensemble learning approach is implemented to combine the base classifiers in three stages. During the training phase of the base classifiers, undersampling techniques and a cost-sensitive random forest (CS-RF) are utilized to address the imbalanced data at both the data and algorithmic levels. In the first stage, imbalanced datasets are balanced using random undersampling, particle swarm optimization-based undersampling, and random oversampling. In the second stage, a classifier is constructed for each of the balanced datasets obtained through these sampling techniques. In the third stage, a majority voting method is introduced to aggregate the predicted outputs from the three classifiers. The evaluation results demonstrate that our proposed method significantly enhances the detection of illegitimate users in the imbalanced Twitter dataset. Additionally, we compare our proposed work with existing models, and the predicted results highlight the superiority of our spam detection model over state-of-the-art spam detection models that address the class imbalance problem. The combination of particle swarm optimization-based undersampling and the ensemble learning approach using majority voting results in more accurate spam detection.

Data mining and natural language processing researchers have been working on sentiment analysis for the past decade. Using deep neural networks (DNNs) for sentiment analysis has recently shown promising results. A technique of studying people’s attitudes through emotional sentiment analysis of data generated from various sources such as Twitter, social media reviews, etc. and classifying emotions based on the given data is related to text data generation. Therefore, the proposed study proposes a well-known deep learning technique for facet-based emotional mood classification using text data that can handle a large amount of content. Text data pre-processing uses stemming, segmentation, tokenization, case folding, and removal of stop words, nulls, and special characters. After data pre-processing, three word embedding approaches such as Assimilated N-gram Approach (ANA), Boosted Term Frequency Inverse Document Frequency (BT-IDF) and Enhanced Two-Way Encoder Representation from Transformers (E-BERT) are used to extract relevant features. The extracted features from the three different approaches are concatenated using the Feature Fusion Approach (FFA). The optimal features are selected using the Intensified Hunger Games Search Optimization (I-HGSO) algorithm. Finally, aspect-based sentiment analysis is performed using the Senti-BILSTM (Deep Aspect-EMO SentiNet) autoencoder based on the Hybrid Emotional Aspect Capsule autoencoder. The experiment was built on the yelp reviews dataset, IDMB movie review dataset, Amazon reviews dataset and the Twitter sentiment dataset. A statistical evaluation and comparison of the experimental results are conducted with respect to the accuracy, precision, specificity, the f1-score, recall, and sensitivity. There is a 99.26% accuracy value in the Yelp reviews dataset, a 99.46% accuracy value in the IMDB movie reviews dataset, a 99.26% accuracy value in the Amazon reviews dataset and a 99.93% accuracy value in the Twitter sentiment dataset.

Uninorms combining t-conorms and t-norms on bounded lattices have lately drawn extensive interest. In this article, we propose two ways for constructing uninorms on a bounded lattice with an identity element. They benefit from the appearance of the t-norm (resp. t-conorm) and the closure operator (resp. interior operator) on a bounded lattice. Additionally, we include some illustrative examples to highlight that our procedures differ from others in the literature.