Large language models (LLMs) have achieved remarkable success across a wide range of applications especially when augmented by external knowledge through retrieval-augmented generation (RAG). Despite their widespread adoption, recent studies have shown that LLMs often struggle to perform faithful reasoning when conflicting knowledge is retrieved. However, existing work primarily focuses on conflicts between external knowledge and the parametric knowledge of LLMs, leaving conflicts across external knowledge largely unexplored. Meanwhile, modern RAG systems increasingly emphasize the integration of unstructured text and (semi-)structured data like knowledge graphs (KGs) to improve knowledge completeness and reasoning faithfulness. To address this gap, we introduce ConflictQA, a novel benchmark that systematically instantiates conflicts between textual evidence and KG evidence. Extensive evaluations across representative LLMs reveal that, facing such cross-source conflicts, LLMs often fail to identify reliable evidence for correct reasoning. Instead, LLMs become more sensitive to prompting choices and tend to rely exclusively on either KG or textual evidence, resulting in incorrect responses. Based on these findings, we further propose XoT, a two-stage explanation-based thinking framework tailored for reasoning over heterogeneous conflicting evidence, and verify its effectiveness with extensive experiments.

Large language models (LLMs) have emerged as a promising paradigm for recommender systems, due to their powerful capabilities in global knowledge integration and reasoning. However, LLMs are inherently prone to confirmation bias -- the tendency to favor information that reinforces users' existing views -- which leads to an overemphasis on previously shown viewpoints and ignores diverse user beliefs for recommendations. To address this issue, in this paper, we propose SCoTRec, a social chain-of-thought reasoning framework for recommendation. SCoTRec first constructs sentiment-aware user profiles by extracting sentiment terms from user reviews. It then incorporates users' social sentiment information into the social chain-of-thought reasoning units to improve recommendations. In particular, we categorize the social chain-of-thought into sentiment-based pathways and apply human evaluation operations -- backtracking, discarding, retaining, and aggregating -- to simulate nuanced sentiment cognition and interpersonal influence, effectively alleviating confirmation bias. Extensive experiments on four benchmark datasets demonstrate the effectiveness of SCoTRec in alleviating confirmation bias and improving recommendations.

Deep learning has made remarkable progress in medical image analysis, yet its black-box nature still limits interpretability and clinician trust. Concept-based modeling offers a promising direction for explainable AI by integrating human-understandable concepts. However, existing approaches typically rely on global concept annotations and infer diagnosis based solely on the presence or absence of individual concepts. This oversimplified paradigm ignores the rich relationships among concepts and their causal influence on disease outcomes. To overcome these limitations, we propose the Concept Relationship Embedding Model (CREM) for interpretable medical diagnosis. CREM mirrors coarse-to-fine clinical reasoning by first extracting fine-grained subregional concepts, then explicitly encoding their relationships as a concept interaction graph, and finally performing causal inference between concepts and diagnoses to enable reliable and transparent diagnostic predictions. We evaluate CREM on four public medical imaging benchmarks, where it achieves state-of-the-art performance on both concept recognition and disease classification tasks, while exhibiting improved robustness, label efficiency, and interpretability. Furthermore, we deploy CREM as an interactive web-based demo that allows clinicians to visualize concept activations, trace diagnostic reasoning paths, and iteratively refine concept cues, facilitating effective human-in-the-loop decision-making.

Most self-supervised graph learning studies typically follow an offline training paradigm, assuming that all data are readily available.This assumption, however, is not always tenable in real-world scenarios as many graph data are generated continuously.Although several continual graph learning models have emerged and achieved empirical success, they almost all rely on external supervision, making it difficult to adapt to applications with a large amount of unlabeled data from the wild.To be honest, research on self-supervised continual graph learning is still surprisingly in its infancy.Therefore, we select several well-known self-supervised graph embedding models as representatives and explore whether they are resistant to catastrophic forgetting in a continual learning setting.Empirical studies find that self-supervised representation models may be potentially better continual learners than supervised counterparts.Driven by this advantage, we propose a self-supervised continual graph representation learning framework based on adaptive spaced replay on node proxies, named Trace.Inspired by the Complementary Learning System theory, Trace employs a dual-system architecture to simulate the functionality and cooperation of the hippocampus and neocortex in the brain.Among them, the fastlearning system efficiently encodes the current input graph to acquire new knowledge and adaptively extracts node proxies from it as important knowledge cached into the memory through progressive clustering.Drawing inspiration from the Ebbinghaus forgetting curve, the slow-learning system implements adaptive spaced replay based on the memory retention rate of each preceding task instead of the widely used consecutive replay scheme for promising flexibility and efficiency.Experiments under task-incremental and class-incremental learning settings on multiple datasets corroborate

Diagram question answering (DQA) is a challenging task that requires models to combine with domain-specific knowledge and reason over the diagrams to answer questions. Multimodal Large Language Models (MLLMs) have recently made notable strides in combining textual and visual information, emerging as a promising solution for addressing the DQA task. However, they still encounter challenges in deliberate multimodal reasoning over the fine-grained visual details of content-rich and knowledge-grounded diagrams. The tight interweaving of visual and textual reasoning for MLLMs is also susceptible to hallucinations. To overcome these limitations, we propose a Glance-then-Focus Multimodal Reasoning framework named GlFoMR for DQA, which features a flexible architecture for comprehensive visual and text interaction. Firstly, the diagram is parsed into a hierarchical structure spanning different granularities including isolated single-object, object-group, and whole-diagram. Subsequently, the Glance-Plan and Focus-Reason stages collaborate to decouple the complex reasoning process. Glance-Plan first generates a preliminary plan by glancing at the multimodal context, specifying sub-goals related to knowledge extraction, visual perception, and visual reasoning. Based on these sub-goals, Focus-Reason further integrates domain-specific knowledge and visual details to enable more deliberate reasoning. The parsed multi-granularity diagram information is seamlessly incorporated into the corresponding sub-goal achievement process, enhancing the perception and reasoning capabilities of MLLMs for better DQA performance. Extensive experimental results on four DQA datasets demonstrate that GlFoMR achieves substantial improvements, showcasing its potential to advance the development of multimodal reasoning.

Continual knowledge graph embedding (CKGE) has gained wide attention for managing dynamic knowledge graphs (KGs), which are continuously updated with new facts. Unlike traditional methods designed for static KGs, CKGE enables incremental updates to KG embeddings to accommodate new facts while retaining previously learned knowledge. Despite these advancements, current CKGE studies and benchmarks primarily focus on handling the increasing scale of data while overlooking changes in graph patterns. These changes, altering the graph structure of KGs, are referred to as pattern shifts in this paper. Pattern shifts frequently arise as new facts are added, introducing significant challenges to the stability and adaptability of CKGE methods. To address this gap, we introduce a suite of novel and challenging benchmarks, called PS-CKGE, specifically designed to evaluate CKGE methods under pattern shifts, where logic rules are utilized to capture and manage structural changes in dynamic KGs. Through these benchmarks, we comprehensively evaluate current CKGE methods in terms of their overall performance, resistance to catastrophic forgetting, and adaptability to new knowledge. The results show that pattern shifts not only exacerbate their risk of catastrophic forgetting but also impair their adaptability, usually with greater performance degradation over triples associated with more significant changes.

Logical reasoning consistently plays a fundamental and significant role in the domains of knowledge engineering and artificial intelligence. Recently, Large Language Models (LLMs) have emerged as a noteworthy innovation in natural language processing (NLP). However, the question of whether LLMs can effectively address the task of logical reasoning, which requires gradual cognitive inference similar to human intelligence, remains unanswered. To this end, we aim to bridge this gap and provide comprehensive evaluations in this paper. First, to offer systematic evaluations, we select fifteen typical logical reasoning datasets and organize them into deductive, inductive, abductive and mixed-form reasoning settings. Considering the comprehensiveness of evaluations, we include 3 early-era representative LLMs and 4 trending LLMs. Second, different from previous evaluations relying only on simple metrics (e.g.,accuracy), we propose fine-level evaluations in objective and subjective manners, covering both answers and explanations, includinganswer correctness,explain correctness,explain completenessandexplain redundancy. Additionally, to uncover the logical flaws of LLMs, problematic cases will be attributed to five error types from two dimensions, i.e.,evidence selection processandreasoning process. Third, to avoid the influences of knowledge bias and concentrate purely on benchmarking the logical reasoning capability of LLMs, we propose a new dataset with neutral content. Based on the in-depth evaluations, this paper finally forms a general evaluation scheme of logical reasoning capability from six dimensions (i.e.,Correct,Rigorous,Self-aware,Active,OrientedandNo hallucination). It reflects the pros and cons of LLMs and gives guiding directions for future works.

Embedding symbolic logical formulas into a low-dimensional continuous space provides an effective way for the Neural-Symbolic system. However, current studies are all constrained by the syntactic structure modeling and fail to preserve intrinsic semantics. To this end, we propose a novel model of Contrastive Graph Representations (ConGR) for logical formulas embedding. Firstly, it introduces a densely connected graph convolutional network (GCN) with an attention mechanism to process syntax parsing graphs of formulas. Secondly, the contrastive instances for each anchor formula are generated by the transformation under the guidance of logical properties. Two types of contrast, global-local and global-global, are carried out to refine formula embeddings with semantic information. Extensive experiments demonstrate that ConGR obtains superior performance against state-of-the-art baselines.

Currently, the increasing information overload on Massive Open Online Courses(MOOCs) inhibits the appropriate choice of learning objects by learners, leading to low efficiency and high dropout rates. However, in MOOC platforms, recommendation network structures that can selectively extract implicit features such as heterogeneous learning preference and knowledge organization of learning objects are still not comprehensively studied. To this end, we propose a learning object recommendation model namely ACGCN based on heterogeneous learning behavior and knowledge graph. By introducing an attention mechanism, information is amplified when updating the representation of the heterogeneous graph, which eliminates the impact of noise and improves the robustness of ACGCN. Experimental results using a real-world dataset revealed that our proposed model has the best performance compared to those of several existing baselines.

Knowledge graph embedding (KGE) methods have achieved great success in handling various knowledge graph (KG) downstream tasks. However, KGE methods may learn biased representations on low-quality KGs that are prevalent in the real world. Some recent studies propose adversarial attacks to investigate the vulnerabilities of KGE methods, but their attackers are target-oriented with the KGE method and the target triples to predict are given in advance, which lacks practicability. In this work, we explore untargeted attacks with the aim of reducing the global performances of KGE methods over a set of unknown test triples and conducting systematic analyses on KGE robustness. Considering logic rules can effectively summarize the global structure of a KG, we develop rule-based attack strategies to enhance the attack efficiency. In particular, we consider adversarial deletion which learns rules, applying the rules to score triple importance and delete important triples, and adversarial addition which corrupts the learned rules and applies them for negative triples as perturbations. Extensive experiments on two datasets over three representative classes of KGE methods demonstrate the effectiveness of our proposed untargeted attacks in diminishing the link prediction results. And we also find that different KGE methods exhibit different robustness to untargeted attacks. For example, the robustness of methods engaged with graph neural networks and logic rules depends on the density of the graph. But rule-based methods like NCRL are easily affected by adversarial addition attacks to capture negative rules.

Modeling the dynamics into graph neural networks (GNNs) contributes to the understanding of evolution in dynamic graphs, which helps optimize temporal-spatial representations for real-world dynamic network problems. Empirically, dynamic GNN embedding requires additional temporal encoders, which inevitably introduces additional learning parameters to make dynamic GNNs oversized and inefficient. Furthermore, previous dynamic GNN models are under the same fixed temporal term, which causes the short-temporal optimum. To address these issues, we propose the WinGNN framework to model dynamic graphs, which is realized by a simple GNN model with the meta-learning strategy and a novel mechanism of random gradient aggregation. WinGNN calculates the frame-wise loss of the current snapshot and passes the loss gradient to the next to model graph dynamics without temporal encoders. Then it introduces the randomized sliding-window to acquire the window-aware gradienton consecutive snapshots, and the calculated two types of gradient are aggregated to update the GNN, thereby reducing the parameter size and improving the robustness. Experiments on six public datasets show the advantage of our WinGNN compared with existing baselines, where it has reached the optimum in twenty-two out of twenty-four performance metrics.

Currently, the non-transparent computing process of deep learning has become a significant reason hindering its further development. The Neural-Symbolic (NS) system formed by integrating logic rules into neural networks has attracted increasing attention owing to its direct interpretability. Embedding symbolic logical formulas into a low-dimensional continuous space provides an effective way for the NS system. However, current studies are all constrained by the modeling ability for its syntactic structure and fail to preserve the intrinsic semantics in embeddings, which causes poor performance on downstream reasoning tasks. To this end, this paper proposes a novel method ofContrastiveGraphRepresentations (ConGR) for logical formulas embedding. First, to improve the modeling ability for the syntactic structure, ConGR introduces a densely connected graph convolutional network (GCN) with an attention mechanism to process syntax parsing graphs of formulas. In this way, discriminative local and global embeddings of formulas are obtained at the syntax level. Second, the contrastive instances (positive or negative) for each anchor formula are generated by the transformation under the guidance of logical properties. To preserve semantic information, two types of contrast, global-local and global-global, are carried out to refine formula embeddings. Extensive experiments demonstrate that ConGR obtains superior performance against state-of-the-art baselines on entailment checking and premise selection datasets.

Few-shot learning (FSL) that acquires new knowledge with little supervision, attracts much attention due to expensive cost of data annotation. Various meta-learning methods have made a great progress for few-shot problem in image and text data. In reality, data samples are not independent but rich in link relations. Large amounts of data exists in the form of graph structure such as citation, social, and biological networks. However, FSL study on graph data is still in its infancy because of the obstacle on extracting meta-knowledge from a meta node classification task. Current research just simply combines the FSL methods experienced in computer vision with node representation models together, but ignores the effect of rich links among support and query nodes in few-shot meta-task. For this issue, we propose a novel Multi-Level Graph Relation Network (MuL-GRN) for the challenging few-shot node classification. MuL-GRN extracts node embeddings through the popular graph neural networks (GNNs). And it includes a relation learning module to mine the deep node relations from three views, namely node-level, global subgraph-level, and local subgraph-level relations. For any two nodes, the node-level relation is computed on their node embeddings, global subgraph-level relation is measured on their subgraph embeddings, and the local subgraph-level relation is mined according to the pairwise node comparison information in their subgraphs. The three-view relation vectors are fused together with an interesting relation fusion module, which measures the importance of relation vector for the current few-shot classification task automatically. Extensive experiments on five real datasets show that MuL-GRN significantly outperforms existing state-of-the-art methods by a large margin.

Massive Open Online Courses (MOOCs) have received unprecedented attention, in which learners can obtain a large number of learning objects anytime and anywhere. However, the increasing information overload on MOOCs inhibits the appropriate choice of learning objects by learners, leading to a low efficiency and high dropout rates in the learning process of this human-computer interaction scenario. E-learning recommendation systems have been studied to present learning objects directly to learners, thereby relieving such problem. However, in MOOC platforms, recommendation network structures which can selectively extract implicit feature such as heterogeneous learning preference and knowledge organization of learning objects are still not comprehensively studied. To this end, we propose a learning object recommendation model based on heterogeneous learning behavior and knowledge graph. To generate a unified representation of each entity and relation, we first propose an Attentive Composition based Graph Convolutional Network (ACGCN). By introducing an attention mechanism, information is amplified when updating the representation of the heterogeneous graph, which eliminates the impact of noise and improves the robustness of the model. Then, a Dense Feature based Operation-Aware Network (DFOAN) is utilized to capture implicit and complex learners’ interactive behaviors, and to further provide a recommendation. Experimental results using two real-world datasets revealed that our proposed model has the best precision, recall, F1, and accuracy scores compared to those of several existing models.

Relation prediction on knowledge graphs (KGs) aims to infer missing valid triples from observed ones. Although this task has been deeply studied, most previous studies are limited to the transductive setting and cannot handle emerging entities. Actually, the inductive setting is closer to real-life scenarios because it allows entities in the testing phase to be unseen during training. However, it is challenging to precisely conduct inductive relation prediction as there exists requirements of entity-independent relation modeling and discrete logical reasoning for interoperability. To this end, we propose a novel model ConGLR to incorporate context graph with logical reasoning. Firstly, the enclosing subgraph w.r.t. target head and tail entities are extracted and initialized by the double radius labeling. And then the context graph involving relational paths, relations and entities is introduced. Secondly, two graph convolutional networks (GCNs) with the information interaction of entities and relations are carried out to process the subgraph and context graph respectively. Considering the influence of different edges and target relations, we introduce edge-aware and relation-aware attention mechanisms for the subgraph GCN. Finally, by treating the relational path as rule body and target relation as rule head, we integrate neural calculating and logical reasoning to obtain inductive scores. And to focus on the specific modeling goals of each module, the stop-gradient is utilized in the information interaction between context graph and subgraph GCNs in the training process. In this way, ConGLR satisfies two inductive requirements at the same time. Extensive experiments demonstrate that ConGLR obtains outstanding performance against state-of-the-art baselines on twelve inductive dataset versions of three common KGs.

Machine reading comprehension has aroused wide concerns, since it explores the potential of model for text understanding. To further equip the machine with the reasoning capability, the challenging task of logical reasoning is proposed. Previous works on logical reasoning have proposed some strategies to extract the logical units from different aspects. However, there still remains a challenge to model the long distance dependency among the logical units. Also, it is demanding to uncover the logical structures of the text and further fuse the discrete logic to the continuous text embedding. To tackle the above issues, we propose an end-to-end model Logiformer which utilizes a two-branch graph transformer network for logical reasoning of text. Firstly, we introduce different extraction strategies to split the text into two sets of logical units, and construct the logical graph and the syntax graph respectively. The logical graph models the causal relations for the logical branch while the syntax graph captures the co-occurrence relations for the syntax branch. Secondly, to model the long distance dependency, the node sequence from each graph is fed into the fully connected graph transformer structures. The two adjacent matrices are viewed as the attention biases for the graph transformer layers, which map the discrete logical structures to the continuous text embedding space. Thirdly, a dynamic gate mechanism and a question-aware self-attention module are introduced before the answer prediction to update the features. The reasoning process provides the interpretability by employing the logical units, which are consistent with human cognition. The experimental results show the superiority of our model, which outperforms the state-of-the-art single model on two logical reasoning benchmarks.

Coreference resolution aims at recognizing different forms in a document which refer to the same entity in the real world. Although many models have been proposed and achieved success, there still exist some challenges. Recent models that use recurrent neural networks to obtain mention representations ignore dependencies between spans and their proceeding distant spans, which will lead to predicted clusters that are locally consistent but globally inconsistent. In addition, these models are trained only by maximizing the marginal likelihood of gold antecedent spans from coreference clusters, which will make some gold mentions undetectable and cause unsatisfactory coreference results. To address these challenges, we propose a neural coreference resolution model. It employs mutual attention to take into account the dependencies between spans and their proceeding spans directly (use attention mechanism to capture global information between spans and their proceeding spans). And our model is trained by jointly optimizing mention clustering and imbalanced mention detection, which enables it to detect more gold mentions in a document to make more accurate coreference decisions. Experimental results on the CoNLL-2012 English dataset show that our model can detect the most gold mentions and achieve the state-of-the-art coreference performance compared with baselines.