Detecting fake reviews can help customers make better purchasing decisions and maintain a positive online business environment. In recent years, pre-trained language models have significantly improved the performance of natural language processing tasks. These models are able to generate different representation vectors for each word in different contexts, thus solving the challenge of multiple meanings of a word, which traditional word vector methods such as Word2Vec cannot solve, and, therefore, better capturing the text’s contextual information. In addition, we consider that reviews generally contain rich opinion and sentiment expressions, while most pre-trained language models, including BERT, lack the consideration of sentiment knowledge in the pre-training stage. Based on the above considerations, we propose a new fake review detection model based on a pre-trained language model and convolutional neural network, which is called BSTC. BSTC considers BERT, SKEP, and TextCNN, where SKEP is a pre-trained language model based on sentiment knowledge enhancement. We conducted a series of experiments on three gold-standard datasets, and the findings illustrate that BSTC outperforms state-of-the-art methods in detecting fake reviews. It achieved the highest accuracy on all three gold-standard datasets—Hotel, Restaurant, and Doctor—with 93.44%, 91.25%, and 92.86%, respectively.

The goal of aspect-based sentiment analysis (ABSA) is to identify the sentiment polarity of specific aspects in a context. Recently, graph neural networks have employed dependent tree syntactic information to assess the link between aspects and contextual words; nevertheless, most of this research has neglected phrases that are insensitive to syntactic analysis and the effect between various aspects in a sentence. In this paper, we propose a dual-channel edge-featured graph attention networks model (AS-EGAT), which builds an aspect syntactic graph by enhancing the contextual syntactic dependency representation of key aspect words and the mutual affective relationship between various aspects in the context and builds a semantic graph through the self-attention mechanism. We use the edge features as a significant factor to determine the weight coefficient of the attention mechanism to efficiently mine the edge features of the graph attention networks model (GAT). As a result, the model can connect important sentiment features of related aspects when dealing with aspects that lack obvious sentiment expressions, pay close attention to important word aspects when dealing with multiple-word aspects, and extract sentiment features from sentences that are not sensitive to syntactic dependency trees by looking at semantic features. Experimental results show that our proposed AS-EGAT model is superior to the current state-of-the-art baselines. Compared with the baseline models of LAP14, REST15, REST16, MAMS, T-shirt, and Television datasets, the accuracy of our AS-EGAT model increased by 0.76%, 0.29%, 0.05%, 0.15%, 0.22%, and 0.38%, respectively. The macro-f1 score increased by 1.16%, 1.16%, 1.23%, 0.37%, 0.53%, and 1.93% respectively.

There is relatively little research on deep learning for anomaly detection within the field of deep learning. Existing deep anomaly detection methods focus on the learning of feature reconstruction, but such methods mainly learn new feature representations, and the new features do not fully reflect the original features, leading to inaccurate anomaly scores; in addition, there is an end-to-end deep anomaly detection algorithm, but the method cannot accurately obtain a reference score that matches the data themselves. In addition, in most practical scenarios, the data are unlabeled, and there exist some datasets with labels, but the confidence and accuracy of the labels are very low, resulting in inaccurate results when put into the model, which makes them often designed for unsupervised learning, and thus in such algorithms, the prior knowledge of known anomalous data is often not used to optimize the anomaly scores. To address the two problems raised above, this paper proposes a new anomaly detection model that learns anomaly scores mainly through a variational deviation network (i.e., not by learning the reconstruction error of new features, distance metrics, or random generation, but by learning the normal distribution of normal data). In this model, we force the anomaly scores to deviate significantly from the normal data by a small amount of anomalous data and a reference score generated by variational self-encoding. The experimental results in multiple classes of data show that the new variational deviation network proposed in this paper has higher accuracy among the mainstream anomaly detection algorithms.

Link prediction is critical to completing the missing links in a network or to predicting the generation of new links according to current network structure information, which is vital for analyzing the evolution of a network, such as the logical architecture construction of MEC (mobile edge computing) routing links of a 5G/6G access network. Link prediction can provide throughput guidance for MEC and select appropriate c nodes through the MEC routing links of 5G/6G access networks. Traditional link prediction algorithms are always based on node similarity, which needs predefined similarity functions, is highly hypothetical and can only be applied to specific network structures without generality. To solve this problem, this paper proposes a new efficient link prediction algorithm PLAS (predicting links by analysis subgraph) and its GNN (graph neural network) version PLGAT (predicting links by graph attention networks) based on the target node pair subgraph. In order to automatically learn the graph structure characteristics, the algorithm first extracts the h-hop subgraph of the target node pair, and then predicts whether the target node pair will be linked according to the subgraph. Experiments on eleven real datasets show that our proposed link prediction algorithm is suitable for various network structures and is superior to other link prediction algorithms, especially in some 5G MEC Access networks datasets with higher AUC (area under curve) values.