YPYudai Pan

Yudai Pan

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Yudai Pan has a total of 27 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    TrajLM: Transferable Mobility Pattern Distillation for Unified Trajectory Learning in Sparse Road Networks

    by Xiaolin Han, Jianqiang Gao, Yuting Cui, Chenhao Ma, Yudai Pan, Zhanhuai Li, Xuequn Shang

    Spatio-temporal trajectory data mining plays a pivotal role in understanding and predicting object movements across space and time, impacting crucial location-based applications such as urban traffic optimization, human mobility analysis, and autonomous vehicle navigation. Although traditional methods effectively capture local trajectory patterns, they struggle with the semantic understanding required for robust representation of diverse traffic scenarios due to their limited ability to incorporate global contextual semantics. Recent advances in large language models (LLMs) have shown promise in capturing semantic relationships from extensive data; however, they face difficulties adapting their powerful semantic learning capabilities to sparse regions where only a few trajectories are traversed because of uneven geographical coverage. To bridge this gap, we propose TrajLM , a novel model that distills transferable mobility patterns in sparse road networks for unified trajectory representation via large language models. Our approach facilitates deep collaboration between dense and sparse trajectory regions, leveraging mutual knowledge distillation to mitigate the sparsity challenge and enhance semantic representation. Moreover, by incorporating spatio-temporal dynamics, our model enables LLMs to learn time-varying movement representations, thereby improving mobility understanding. Extensive experiments demonstrate that our model significantly outperforms state-of-the-art competitors across three downstream tasks, achieving an average improvement of 29.51%. Implementation details and code are available at https://anonymous.4open.science/r/TrajLM-6343.

  • Open Access

    Rethinking Continual Knowledge Graph Embedding: Benchmarks and Analysis

    by Tianzhe Zhao, Jiaoyan Chen, Yanchi Ru, Qika Lin, Yuxia Geng, Haiping Zhu, Yudai Pan, Jun Liu


  • Open Access

    A Symbolic Rule Integration Framework with Logic Transformer for Inductive Relation Prediction

    by Yudai Pan, Jun Liu, Tianzhe Zhao, Lingling Zhang, Yun Lin, Jin Song Dong

    Relation prediction in knowledge graphs (KGs) aims at predicting missing relations in incomplete triples, whereas the dominant paradigm by KG embeddings has a limitation to predict the relation between unseen entities. This situation is called an inductive setting, which is more common in the real-world scenario. To handle this issue, implicit symbolic rules have shown great potential in capturing the inductive capability. However, it is still challenging to obtain precise representations of logic rules from KGs. The argument variability and predicate non-commutativity in symbolic rule integration make the modeling of component symbols difficult. To this end, we propose a novel inductive relation prediction model named SymRITa with a logic transformer integrating rules. SymRITa firstly extracts the subgraph, whose embeddings are captured by a graph network. Meanwhile, symbolic rule graphs in the subgraph can be generated. Then, the symbolic rules are modeled by a proposed logic transformer. Specifically, the input format based on the subgraph-based embeddings is to focus on the argument variability in symbolic rules. In addition, a conjunction attention mechanism in the logic transformer can resolve predicate non-commutativity in the symbolic rule integration process. Finally, the subgraph-based and symbol-based embeddings obtained from the previous steps are combined for the training regime, and prediction results as well as rules explaining the reasoning process are explicitly output. Extensive experiments on twelve inductive datasets show that SymRITa achieves outstanding effectiveness compared to state-of-the-art inductive baselines. Moreover, the logic rules with corresponding confidences provide an interpretable paradigm.


  • Context-Aware Commonsense Knowledge Graph Reasoning With Path-Guided Explanations

    by Yudai Pan, Jun Liu, Tianzhe Zhao, Lingling Zhang, Qianying Wang

    Commonsense knowledge graphs (CKGs) store massive commonsense knowledge as triples whose nodes consist of free-form texts. CKG reasoning aims to predict missing nodes in incomplete commonsense triples, which is challenging as it requires more accurate embeddings for reasoning. Compared to conventional knowledge graphs (KGs), CKGs have deficient structural information due to their sparsity and contain nodes indistinguishable due to the conceptual diversity. These issues limit the performance of previous reasoning methods, because they face difficulties obtaining precise CKG representations. To address these issues, we propose a context-aware CKG reasoning framework with path-guided explanations, named CoRPe. First, CoRPe constructs context sentences based on the target commonsense triple using designed templates. The context captures reasoning paths instantiated from the first-order logic. Second, to improve CKG representations, CoRPe injects context semantics and employs a context-augmented tuning strategy on a pre-trained language model (PLM) via a synergistic optimization. Finally, CoRPe embeds structural information using a graph convolutional network (GCN) and associates the textual semantics for joint scoring. Extensive experiments on two CKGs show that CoRPe outperforms state-of-the-art KG and CKG reasoning baselines in terms of embedding and reasoning performance. Furthermore, the interpretability of CoRPe is reflected in the implicit logic during reasoning.


  • Contrastive Graph Representations for Logical Formulas Embedding

    by Qika Lin, Jun Liu, Lingling Zhang, Yudai Pan, Xin Hu, Fangzhi Xu, Hongwei Zeng

    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 of Con trastive G raph R epresentations (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.


  • Open Access

    Learning First-Order Rules with Relational Path Contrast for Inductive Relation Reasoning

    by Yudai Pan, Jun Liu, Lingling Zhang, Xin Hu, Tianzhe Zhao, Qika Lin

    Relation reasoning in knowledge graphs (KGs) aims at predicting missing relations in incomplete triples, whereas the dominant paradigm is learning the embeddings of relations and entities, which is limited to a transductive setting and has restriction on processing unseen entities in an inductive situation. Previous inductive methods are scalable and consume less resource. They utilize the structure of entities and triples in subgraphs to own inductive ability. However, in order to obtain better reasoning results, the model should acquire entity-independent relational semantics in latent rules and solve the deficient supervision caused by scarcity of rules in subgraphs. To address these issues, we propose a novel graph convolutional network (GCN)-based approach for interpretable inductive reasoning with relational path contrast, named RPC-IR. RPC-IR firstly extracts relational paths between two entities and learns representations of them, and then innovatively introduces a contrastive strategy by constructing positive and negative relational paths. A joint training strategy considering both supervised and contrastive information is also proposed. Comprehensive experiments on three inductive datasets show that RPC-IR achieves outstanding performance comparing with the latest inductive reasoning methods and could explicitly represent logical rules for interpretability.


  • Fine-Grained 3D-Attention Prototypes for Few-Shot Learning

    by Xin Hu, Jun Liu, Jie Ma, Yudai Pan, Lingling Zhang

    In the real world, a limited number of labeled finely grained images per class can hardly represent the class distribution effectively. Due to the more subtle visual differences in fine-grained images than simple images with obvious objects, that is, there exist smaller interclass and larger intraclass variations. To solve these issues, we propose an end-to-end attention-based model for fine-grained few-shot image classification (AFG) with the recent episode training strategy. It is composed mainly of a feature learning module, an image reconstruction module, and a label distribution module. The feature learning module mainly devises a 3D-Attention mechanism, which considers both the spatial positions and different channel attentions of the image features, in order to learn more discriminative local features to better represent the class distribution. The image reconstruction module calculates the mappings between local features and the original images. It is constrained by a designed loss function as auxiliary supervised information, so that the learning of each local feature does not need extra annotations. The label distribution module is used to predict the label distribution of a given unlabeled sample, and we use the local features to represent the image features for classification. By conducting comprehensive experiments on Mini-ImageNet and three fine-grained data sets, we demonstrate that the proposed model achieves superior performance over the competitors. In the real world, a limited number of labeled finely grained images per class can hardly represent the class distribution effectively. Due to the more subtle visual differences in fine-grained images than simple images with obvious objects, that is, there exist smaller interclass and larger intraclass variations. To solve these issues, we propose an end-to-end attention-based model for fine-grained few-shot image classification (AFG) with the recent episode training strategy. It is composed mainly of a feature learning module, an image reconstruction module, and a label distribution module. The feature learning module mainly devises a 3D-Attention mechanism, which considers both the spatial positions and different channel attentions of the image features, in order to learn more discriminative local features to better represent the class distribution. The image reconstruction module calculates the mappings between local features and the original images. It is constrained by a designed loss function as auxiliary supervised information, so that the learning of each local feature does not need extra annotations. The label distribution module is used to predict the label distribution of a given unlabeled sample, and we use the local features to represent the image features for classification. By conducting comprehensive experiments on Mini-ImageNet and three fine-grained data sets, we demonstrate that the proposed model achieves superior performance over the competitors. In the real world, a limited number of labeled finely grained images per class can hardly represent the class distribution effectively. Due to the more subtle visual differences in fine-grained images than simple images with obvious objects, that is, there exist smaller interclass and larger intraclass variations. To solve these issues, we propose an end-to-end attention-based model for fine-grained few-shot image classification (AFG) with the recent episode training strategy. It is composed mainly of a feature learning module, an image reconstruction module, and a label distribution module. The feature learning module mainly devises a 3D-Attention mechanism, which considers both the spatial positions and different channel attentions of the image features, in order to learn more discriminative local features to better represent the class distribution. The image reconstruction module calculates the mappings between local features and the original images. It is constrained by a designed loss function as auxiliary supervised information, so that the learning of each local feature does not need extra annotations. The label distribution module is used to predict the label distribution of a given unlabeled sample, and we use the local features to represent the image features for classification. By conducting comprehensive experiments on Mini-ImageNet and three fine-grained data sets, we demonstrate that the proposed model achieves superior performance over the competitors.


  • Open Access

    Collective Intelligence: A Taxonomy and Survey

    by Feijuan He, Yudai Pan, Qika Lin, Xianglin Miao, Zhouguo Chen

    Collective intelligence (CI) refers to the intelligence that emerges at the macro-level of a collection and transcends that of the individuals. CI is a continuously popular research topic that is studied by researchers in different areas, such as sociology, economics, biology, and artificial intelligence. In this survey, we summarize the works of CI in various fields. First, according to the existence of interactions between individuals and the feedback mechanism in the aggregation process, we establish CI taxonomy that includes three paradigms: isolation, collaboration and feedback. We then conduct statistical literature analysis to explain the differences among three paradigms and their development in recent years. Second, we elaborate the types of CI under each paradigm and discuss the generation mechanism or theoretical basis of the different types of CI. Third, we describe certain CI-related applications in 2019, which can be appropriately categorized by our proposed taxonomy. Finally, we summarize the future research directions of CI under each paradigm. We hope that this survey helps researchers understand the current conditions of CI and clears the directions of future research.