Nowadays, numerous online platforms can be described as multi-modal heterogeneous networks (MMHNs), such as Douban's movie networks and Amazon's product review networks. Accurately categorizing nodes within these networks is crucial for analyzing the corresponding entities, which requires effective representation learning on nodes. However, existing multi-modal fusion methods often adopt either early fusion strategies which may lose the unique characteristics of individual modalities, or late fusion approaches overlooking the cross-modal guidance in GNN-based information propagation. In this paper, we propose a novel model for node classification in MMHNs, named Heterogeneous Graph Neural Network with Inter-Modal Attention (HGNN-IMA). It learns node representations by capturing the mutual influence of multiple modalities during the information propagation process, within the framework of heterogeneous graph transformer. Specifically, a nested inter-modal attention mechanism is integrated into the inter-node attention to achieve adaptive multi-modal fusion, and modality alignment is also taken into account to encourage the propagation among nodes with consistent similarities across all modalities. Moreover, an attention loss is augmented to mitigate the impact of missing modalities. Extensive experiments validate the superiority of the model in the node classification task, providing an innovative view to handle multi-modal data, especially when accompanied with network structures. The full version including Appendix is available at http://arxiv.org/abs/2505.07895.

Recently, rule-based time series classification models are widely used in safety-critical scenarios demanding strong interpretability, through providing explicit and rigorous rationales. However, due to the discrepancy between discrete rules and continuous neural networks, the generated rules are not completely consistent with the actual decision-making process, rendering users hesitant to trust the model. Additionally, existing methods learn the role (weight) of each time point independently, which is not in line with human understanding and the consecutive nature of temporal properties. In this paper, we propose a novel neuro-symbolic model named TemporalRule, aiming to automatically learn Signal Temporal Logic (STL) rules for interpretable time series classification. Our model directly optimizes the neural networks representing discrete rules via gradient grafting, producing reliable rules that can exactly determine the classification results. Notably, a temporal logical layer is designed to simulate expressive temporal operators, including one 2-predicate operator (Until) and two 2-level composite operators (EA and AE), through adaptively learning the time bounds of involved intervals. That creates consistent weights for consecutive time points, enhancing the intuitiveness of derived rules. Extensive experiments on diverse real-world datasets in safety-critical domains show that TemporalRule achieves superior and stable classification accuracy compared to state-of-the-art rule-based approaches, and the learned rules can precisely and concisely explicate the classification results, leading to a trustable model.

Graphs constructed from real-world scenarios are often incomplete due to privacy restrictions or resource limitations, posing significant challenges for node classification, especially when labeled data are scarce. In many scenarios of incomplete graphs, the real node degrees, such as the number of followers in social networks or publications' references in citation networks, are easily accessible and informative, which could indicate the degree of incompleteness. However, most of existing researches of incomplete graphs focus on edge completion, but ignore the node completion with known node degrees. In this paper, we propose a new few-shot node classification problem on incomplete graphs with real node degrees. To deal with node completion, edge completion and label completion of this problem, we develop an effective Large Language Models (LLMs) empowered Graph Convolutional Network (GCN) model utilizing the real node Degrees, namely LLMDGCN. First, we leverage LLMs to initially fill in the missing nodes and labels. Next, we design an edge prediction module that employs the real node degrees and inter-category probability matrix to recover the missing edges for each node. We then iteratively train the GCN and the edge prediction module. The GCN generates pseudo labels, which the edge prediction module uses to restore edges, and these edges are fed back into the GCN to improve accuracy. Extensive experiments on four benchmark datasets demonstrate the effectiveness and robustness of our proposed method for the few-shot node classification on incomplete graphs with real node degrees.

Weakly supervised text classification (WSTC), also called zero-shot or dataless text classification, has attracted increasing attention due to its applicability in classifying a mass of texts within the dynamic and open Web environment, since it requires only a limited set of seed words (label names) for each category instead of labeled data. With the help of recently popular prompting Pre-trained Language Models (PLMs), many studies leveraged manually crafted and/or automatically identified verbalizers to estimate the likelihood of categories, but they failed to differentiate the effects of these category-indicative words, let alone capture their correlations and realize adaptive adjustments according to the unlabeled corpus. In this paper, in order to let the PLM effectively understand each category, we at first propose a novel form of rule-based knowledge using logical expressions to characterize the meanings of categories. Then, we develop a prompting PLM-based approach named RulePrompt for the WSTC task, consisting of a rule mining module and a rule-enhanced pseudo label generation module, plus a self-supervised fine-tuning module to make the PLM align with this task. Within this framework, the inaccurate pseudo labels assigned to texts and the imprecise logical rules associated with categories mutually enhance each other in an alternative manner. That establishes a self-iterative closed loop of knowledge (rule) acquisition and utilization, with seed words serving as the starting point. Extensive experiments validate the effectiveness and robustness of our approach, which markedly outperforms state-of-the-art weakly supervised methods. What is more, our approach yields interpretable category rules, proving its advantage in disambiguating easily-confused categories.

Anomaly detection, particularly the detection of anomalous behaviors in dynamic and heterogeneous social networks, is becoming more and more crucial in real life. Traditional rule-based and feature-based methods cannot well capture the structural and temporal patterns of ever-changing user behaviors. Moreover, most of the existing works based on network embedding either rely on discretized snapshots, which have ignored accurate temporal relations among user behaviors and weakened the impact of new edges, or fail to utilize dynamic and heterogeneous information simultaneously to distinguish varying effects of new edges on existing nodes. In this paper, we propose an end-to-end continuous-time model, named Temporal Heterogeneous Graph Neural Network (THGNN), to detect anomalous behaviors (edges) in dynamic heterogeneous social networks. Specifically, the model constantly updates node embeddings by propagating the information of a new edge to its source and target nodes as well as their neighbors. In this process, heterogeneous encoders are employed to handle different types of nodes and edges. What is more, a novel dual-level distributive attention mechanism is designed to allocate the influence degree of a currently interacting node to its multiple neighbors, considering the combined effect of edge type and time interval information. That can be regarded as an extension of the classical aggregative attention mechanism in the opposite direction. Extensive experiments on four real-world datasets demonstrate that THGNN outperforms all the baselines on the task of anomalous edge detection, achieving an average AUC gain of 6% across all datasets.

Continual text classification is an important research direction in Web mining. Existing works are limited to supervised approaches relying on abundant labeled data, but in the open and dynamic environment of Internet, involving constant semantic change of known topics and the appearance of unknown topics, text annotations are hard to access in time for each period. That calls for the technique of weakly supervised text classification (WSTC), which requires just seed words for each category and has succeed in static text classification tasks. However, there are still no studies of applying WSTC methods in a continual learning paradigm to actually accommodate the open and evolving Internet. In this paper, we tackle this problem for the first time and propose a framework, named Continual Learning for Weakly Supervised Text Classification (CL-WSTC), which can take any WSTC method as base model. It consists of two modules, classification decision with delay and seed word updating. In the former, the probability threshold for each category in each period is adaptively learned to determine the acceptance/rejection of texts. In the latter, with candidate words output by the base model, seed words are added and deleted via reinforcement learning with immediate rewards, according to an empirically certified unsupervised measure. Extensive experiments show that our approach has strong universality and can achieve a better trade-off between classification accuracy and decision timeliness compared to non-continual counterparts, with intuitively interpretable updating of seed words.

Relation classification (RC) is an important task in knowledge extraction from texts, while data-driven approaches, although achieving high performance, heavily rely on a large amount of annotated training data. Recently, many few-shot RC models have been proposed and yielded promising results in general domain datasets, but when adapting to a specific domain, such as medicine, the performance drops dramatically. In this paper, we propose a Knowledge-Enhanced Few-shot RC model for the Domain Adaptation task (KEFDA), which incorporates general and domain-specific knowledge graphs (KGs) to the RC model to improve its domain adaptability. With the help of concept-level KGs, the model can better understand the semantics of texts and easily summarize the global semantics of relation types from only a few instances. To be more important, as a kind of meta-information, the manner of utilizing KGs can be transferred from existing tasks to new tasks, even across domains. Specifically, we design a knowledge-enhanced prototypical network to conduct instance matching, and a relation-meta learning network for implicit relation matching. The two scoring functions are combined to infer the relation type of a new instance. Experimental results on the Domain Adaptation Challenge in the FewRel 2.0 benchmark demonstrate that our approach significantly outperforms the state-of-the-art models (by 6.63% on average).

Dataless text classification has attracted increasing attentions recently. It only needs very few seed words of each category to classify documents, which is much cheaper than supervised text classification that requires massive labeling efforts. However, most of existing models pay attention to long texts, but get unsatisfactory performance on short texts, which have become increasingly popular on the Internet. In this paper, we at first propose a novel model named Seeded Biterm Topic Model (SeedBTM) extending BTM to solve the problem of dataless short text classification with seed words. It takes advantage of both word co-occurrence information in the topic model and category-word similarity from widely used word embeddings as the prior topic-in-set knowledge. Moreover, with the same approach, we also propose Seeded Twitter Biterm Topic Model (SeedTBTM), which extends Twitter-BTM and utilizes additional user information to achieve higher classification accuracy. Experimental results on five real short-text datasets show that our models outperform the state-of-the-art methods, and especially perform well when the categories are overlapping and interrelated.

Textual documents created and distributed on the Internet are ever changing in various forms. Most of existing works are devoted to topic modeling and the evolution of individual topics, while sequential relations of topics in successive documents published by a specific user are ignored. In this paper, in order to characterize and detect personalized and abnormal behaviors of Internet users, we propose Sequential Topic Patterns (STPs) and formulate the problem of mining User-aware Rare Sequential Topic Patterns (URSTPs) in document streams on the Internet. They are rare on the whole but relatively frequent for specific users, so can be applied in many real-life scenarios, such as real-time monitoring on abnormal user behaviors. We present a group of algorithms to solve this innovative mining problem through three phases: preprocessing to extract probabilistic topics and identify sessions for different users, generating all the STP candidates with (expected) support values for each user by pattern-growth, and selecting URSTPs by making user-aware rarity analysis on derived STPs. Experiments on both real (Twitter) and synthetic datasets show that our approach can indeed discover special users and interpretable URSTPs effectively and efficiently, which significantly reflect users' characteristics.

Plain text documents created and distributed on the Internet are ever changing in various forms. Mining topics of these documents has significant applications in many domains. Most of the literature is devoted to topic modeling, while sequential patterns of topics in document streams are ignored. Moreover, traditional sequential pattern mining algorithms mainly focused on frequent patterns for deterministic data sets, and thus not suitable for document streams with topic uncertainty and rare patterns. In this paper, we formulate and handle the mining problem of rare Sequential Topic Patterns (STPs) for Internet document streams, which are rare on the whole but relatively often for specific users, so also interesting. Since this type of rare STPs reflects users’ specific behaviors, our work can be applied in many fields, such as personalized context-aware recommendation and real-time monitoring on abnormal user behaviors on the Internet. We propose a novel approach to discovering user-related rare STPs based on the temporal and probabilistic information of concerned topics. After extracting topics from documents by LDA and sorting the document stream into sessions for different users during different time periods, the proposed algorithms discover rare STPs by (1) mining STP candidates for each user through an efficient algorithm based on pattern-growth, and (2) generating user-related rare STPs by pattern rarity analysis. Experiments on both synthetic and real data sets show that our approach can discover interesting rare STPs very effectively and efficiently.

Fraenzle and Hansen investigated the model-checking problem of the subset of Duration Calculus without individual variables and quantifications w.r.t. some approximation semantics by reduction to the decision problem of Presburger Arithmetic, thus obtained a model-checking algorithm with 4-fold exponential complexity [6,7]. As an alternative, inspired by their work, we consider the bounded model-checking problem of the subset in the context of the standard discrete-time semantics in this paper. Based on our previous work [20], we reduce this problem to the reachability problem of timed automata. The complexity of our approach is singly exponential in the size of formulas and quadratic in the number of states of models. We implement our approach using UPPAAL and demonstrate its efficiency by some examples.