Jianliang Gao

Tairan Huang, Yili Wang, Qiutong Li et al.

Graph fraud detection has garnered significant attention as Graph Neural Networks (GNNs) have proven effective in modeling complex relationships within multimodal data. However, existing graph fraud detection methods typically use preprocessed node embeddings and predefined graph structures to reveal fraudsters, which ignore the rich semantic cues contained in raw textual information. Although Large Language Models (LLMs) exhibit powerful capabilities in processing textual information, it remains a significant challenge to perform multimodal fusion of processed textual embeddings with graph structures. In this paper, we propose a \textbf{M}ulti-level \textbf{L}LM \textbf{E}nhanced Graph Fraud \textbf{D}etection framework called MLED. In MLED, we utilize LLMs to extract external knowledge from textual information to enhance graph fraud detection methods. To integrate LLMs with graph structure information and enhance the ability to distinguish fraudsters, we design a multi-level LLM enhanced framework including type-level enhancer and relation-level enhancer. One is to enhance the difference between the fraudsters and the benign entities, the other is to enhance the importance of the fraudsters in different relations. The experiments on four real-world datasets show that MLED achieves state-of-the-art performance in graph fraud detection as a generalized framework that can be applied to existing methods.

Yili Wang, Tairan Huang, Changlong He et al.

Heterogeneous temporal graphs (HTGs) are ubiquitous data structures in the real world. Recently, to enhance representation learning on HTGs, numerous attention-based neural networks have been proposed. Despite these successes, existing methods rely on a decoupled temporal and spatial learning paradigm, which weakens interactions of spatio-temporal information and leads to a high model complexity. To bridge this gap, we propose a novel learning paradigm for HTGs called Simple and Efficient Heterogeneous Temporal Graph N}eural Network (SE-HTGNN). Specifically, we innovatively integrate temporal modeling into spatial learning via a novel dynamic attention mechanism, which retains attention information from historical graph snapshots to guide subsequent attention computation, thereby improving the overall discriminative representations learning of HTGs. Additionally, to comprehensively and adaptively understand HTGs, we leverage large language models to prompt SE-HTGNN, enabling the model to capture the implicit properties of node types as prior knowledge. Extensive experiments demonstrate that SE-HTGNN achieves up to 10x speed-up over the state-of-the-art and latest baseline while maintaining the best forecasting accuracy.

Jiamin Chen, Jianliang Gao, Yibo Chen et al.

Graph neural architecture search has received a lot of attention as Graph Neural Networks (GNNs) has been successfully applied on the non-Euclidean data recently. However, exploring all possible GNNs architectures in the huge search space is too time-consuming or impossible for big graph data. In this paper, we propose a parallel graph architecture search (GraphPAS) framework for graph neural networks. In GraphPAS, we explore the search space in parallel by designing a sharing-based evolution learning, which can improve the search efficiency without losing the accuracy. Additionally, architecture information entropy is adopted dynamically for mutation selection probability, which can reduce space exploration. The experimental result shows that GraphPAS outperforms state-of-art models with efficiency and accuracy simultaneously.

Zhao Li, Yixin Liu, Zhen Zhang et al.

Graph neural networks (GNNs) have emerged as effective approaches for graph analysis, especially in the scenario of semi-supervised learning. Despite its success, GNN often suffers from over-smoothing and over-fitting problems, which affects its performance on node classification tasks. We analyze that an alternative method, the label propagation algorithm (LPA), avoids the aforementioned problems thus it is a promising choice for graph semi-supervised learning. Nevertheless, the intrinsic limitations of LPA on feature exploitation and relation modeling make propagating labels become less effective. To overcome these limitations, we introduce a novel framework for graph semi-supervised learning termed as Cyclic Label Propagation (CycProp for abbreviation), which integrates GNNs into the process of label propagation in a cyclic and mutually reinforcing manner to exploit the advantages of both GNNs and LPA. In particular, our proposed CycProp updates the node embeddings learned by GNN module with the augmented information by label propagation, while fine-tunes the weighted graph of label propagation with the help of node embedding in turn. After the model converges, reliably predicted labels and informative node embeddings are obtained with the LPA and GNN modules respectively. Extensive experiments on various real-world datasets are conducted, and the experimental results empirically demonstrate that the proposed CycProp model can achieve relatively significant gains over the state-of-the-art methods.

Zhao Li, Donghui Ding, Pengcheng Zou et al.

The booming online e-commerce platforms demand highly accurate approaches to segment queries that carry the product requirements of consumers. Recent works have shown that the supervised methods, especially those based on deep learning, are attractive for achieving better performance on the problem of query segmentation. However, the lack of labeled data is still a big challenge for training a deep segmentation network, and the problem of Out-of-Vocabulary (OOV) also adversely impacts the performance of query segmentation. Different from query segmentation task in an open domain, e-commerce scenario can provide external documents that are closely related to these queries. Thus, to deal with the two challenges, we employ the idea of distant supervision and design a novel method to find contexts in external documents and extract features from these contexts. In this work, we propose a BiLSTM-CRF based model with an attention module to encode external features, such that external contexts information, which can be utilized naturally and effectively to help query segmentation. Experiments on two datasets show the effectiveness of our approach compared with several kinds of baselines.

Linlin Hou, Ji Zhang, Ou Wu et al.

Literature analysis facilitates researchers to acquire a good understanding of the development of science and technology. The traditional literature analysis focuses largely on the literature metadata such as topics, authors, abstracts, keywords, references, etc., and little attention was paid to the main content of papers. In many scientific domains such as science, computing, engineering, etc., the methods and datasets involved in the scientific papers published in those domains carry important information and are quite useful for domain analysis as well as algorithm and dataset recommendation. In this paper, we propose a novel entity recognition model, called MDER, which is able to effectively extract the method and dataset entities from the main textual content of scientific papers. The model utilizes rule embedding and adopts a parallel structure of CNN and Bi-LSTM with the self-attention mechanism. We evaluate the proposed model on datasets which are constructed from the published papers of four research areas in computer science, i.e., NLP, CV, Data Mining and AI. The experimental results demonstrate that our model performs well in all the four areas and it features a good learning capacity for cross-area learning and recognition. We also conduct experiments to evaluate the effectiveness of different building modules within our model which indicate that the importance of different building modules in collectively contributing to the good entity recognition performance as a whole. The data augmentation experiments on our model demonstrated that data augmentation positively contributes to model training, making our model much more robust in dealing with the scenarios where only small number of training samples are available. We finally apply our model on PAKDD papers published from 2009-2019 to mine insightful results from scientific papers published in a longer time span.

Zheng Chen, Xinli Yu, Chi Zhang et al.

Botnet, a group of coordinated bots, is becoming the main platform of malicious Internet activities like DDOS, click fraud, web scraping, spam/rumor distribution, etc. This paper focuses on design and experiment of a new approach for botnet detection from streaming web server logs, motivated by its wide applicability, real-time protection capability, ease of use and better security of sensitive data. Our algorithm is inspired by a Principal Component Analysis (PCA) to capture correlation in data, and we are first to recognize and adapt Lanczos method to improve the time complexity of PCA-based botnet detection from cubic to sub-cubic, which enables us to more accurately and sensitively detect botnets with sliding time windows rather than fixed time windows. We contribute a generalized online correlation matrix update formula, and a new termination condition for Lanczos iteration for our purpose based on error bound and non-decreasing eigenvalues of symmetric matrices. On our dataset of an ecommerce website logs, experiments show the time cost of Lanczos method with different time windows are consistently only 20% to 25% of PCA.