Graph-based Retrieval-Augmented Generation (GraphRAG) mitigates hallucinations in Large Language Models (LLMs) by grounding them in structured knowledge. However, current GraphRAG methods are constrained by a prevailing build-then-reason paradigm, which relies on a static, pre-constructed Knowledge Graph (KG). This paradigm faces two critical challenges. First, the KG's inherent incompleteness often breaks reasoning paths. Second, the graph’s low signal-to-noise ratio introduces distractor facts, presenting query-relevant but misleading knowledge that disrupts the reasoning process. To address these challenges, we argue for a reason-and-construct paradigm and propose Relink, a framework that dynamically builds a query-specific evidence graph. To tackle incompleteness, Relink instantiates required facts from a latent relation pool derived from the original text corpus, repairing broken paths on the fly. To handle misleading or distractor facts, Relink employs a unified, query-aware evaluation strategy that jointly considers candidates from both the KG and latent relations, selecting those most useful for answering the query rather than relying on their pre-existence. This empowers Relink to actively discard distractor facts and construct the most faithful and precise evidence path for each query. Extensive experiments on five Open-Domain Question Answering benchmarks show that Relink achieves significant average improvements of 5.4% in EM and 5.2% in F1 over leading GraphRAG baselines, demonstrating the superiority of our proposed framework.

Multimodal Entity Alignment (MMEA) aims to identify equivalent entities across different data modalities, enabling structural data integration that in turn improves the performance of various large language model applications. To lift the requirement of labeled seed pairs that are difficult to obtain, recent methods shifted to an unsupervised paradigm using pseudo-alignment seeds. However, unsupervised entity alignment in multimodal settings remains underexplored, mainly because the incorporation of multimodal information often results in imbalanced coverage of pseudo-seeds within the knowledge graph. To overcome this, we propose PSQE (Pseudo-Seed Quality Enhancement) to improve the precision and graph coverage balance of pseudo seeds via multimodal information and clustering-resampling. Theoretical analysis reveals the impact of pseudo seeds on existing contrastive learning-based MMEA models. In particular, pseudo seeds can influence the attraction and the repulsion terms in contrastive learning at once, whereas imbalanced graph coverage causes models to prioritize high-density regions, thereby weakening their learning capability for entities in sparse regions. Experimental results validate our theoretical findings and show that PSQE as a plug-and-play module can improve the performance of baselines by considerable margins.

Cross-modal retrieval (CMR) has advanced considerably by mapping image and text features into a shared embedding space; however, these approaches still face two persistent challenges: (1) semantic sparsity, where discriminative cues are confined to localized regions, making it difficult to identify implicit visual evidence; and (2) ranking uncertainty under semantic ambiguity, where models struggle to maintain the correct retrieval order when candidates share similar contexts. To address these issues, we propose the Dual-Branch Multi-Granularity Network (DBMG) with Structured Contrastive Ranking, which enriches visual semantics by leveraging a multimodal large language model to generate auxiliary descriptions, aligns sparse cues through a dual-branch architecture capturing both global and local interactions, and enforces ranking consistency via a three-stage contrastive objective that progressively optimizes category clustering, instance alignment, and margin-based ranking. Extensive experiments on four standard CMR benchmarks demonstrate that DBMG outperforms 12 strong baselines, achieving an average 15.91% improvement in mAP, establishing a new state-of-the-art. The code is available at https://github.com/DMiC-Lab-HFUT/DBMG.

Cognitive diagnosis (CD) is an important task in the field of intelligent education, aiming to discover the proficiency of students on knowledge concepts with response logs. In applications, different users of the tutoring system demand for a diagnosis of knowledge concepts at different granularities. However, recent methods assume that the concepts are of the same granularity and use explicit correlations between same-granularity concepts to improve the diagnosis performance. If required for diagnosing multi-granularity concepts, these methods will face diminished performance or partial invalidation. To this end, we make the first attempt for multiple-granularity cognitive diagnosis, i.e., diagnosis on coarse- and fine-grained concepts simultaneously. Specifically, in a skillful way, the same-granularity correlations are captured and embedded into concept representations in view of concept semantics and cross-granularity correlations to model the proficiency influence between concepts implicitly. Then, the specific loss for single-granularity diagnosis and the general loss for the consistency of multi-granularity are designed to train the model jointly, achieving multiple-granularity diagnosis. Extensive experiments demonstrate that our method can achieve state-of-the-art accuracy on both coarse- and fine-grained concepts.

Knowledge tracing (KT) refers to the task of analyzing a student’s knowledge mastery over time based on sequential exercise-answering data and has been widely applied in educational settings, such as intelligent tutoring systems. Existing KT models often require manual selection of optimal hyperparameters for the current dataset. However, this manual adaptation process is both time-consuming and resource-intensive, limiting the scalability of KT models in practice. Although automatic architecture search can automate the discovery of optimal hyperparameters and model architectures, it is computationally expensive. This is especially true when the initial configuration is suboptimal, potentially requiring extensive search time to find satisfactory hyperparameters. To address these challenges, we explore the use of evolutionary transfer optimization to efficiently adapt KT models to new data. By integrating automated machine learning techniques with a genetic algorithm-based meta-learning approach, AutoKT transfers optimization knowledge from previously solved tasks to initialize model architectures and hyperparameters for new datasets. Experimental results demonstrate that this initialization can significantly reduce the computational cost of architecture search and hyperparameter tuning.

Entity alignment (EA) aims to find the same entities in different knowledge graphs. Existing EA methods assume the supervised aligned pairs without noise. In applications, noisy pairs lead to degradation of EA performance. To this end, a robust EA method based on knowledge distillation is proposed for noisy pairs. Firstly, the dual-teacher model with online distillation is designed, in which, noise discriminator is performed to improve the noise resistance of teacher models. Secondly, a student model is offline distilled from the dual-teacher model without using the noisy supervised pairs, further enhancing the robustness of student model. In addition, the entity structure is combined with entity representation for alignment inference to alleviate the bias of entity representation in noisy environment. Extensive experiments demonstrate the effectiveness of the proposed method.

Knowledge Graph (KG)-augmented Large Language Models (LLMs) have recently propelled significant advances in complex reasoning tasks, thanks to their broad domain knowledge and contextual awareness. Unfortunately, current methods often assume KGs to be complete, which is impractical given the inherent limitations of KG construction and the potential loss of contextual cues when converting unstructured text into entity-relation triples. In response, this paper proposes the Triple Context Restoration and Query-driven Feedback (TCR-QF) framework, which reconstructs the textual context underlying each triple to mitigate information loss, while dynamically refining the KG structure by iteratively incorporating query-relevant missing knowledge. Experiments on five benchmark question-answering datasets substantiate the effectiveness of TCR-QF in KG and LLM integration, where itachieves a 29.1% improvement in Exact Match and a 15.5% improvement in F1 over its state-of-the-art GraphRAG competitors. The code is publicly available at https://github.com/HFUT-DMiC-Lab/TCR-QF.git.

Relation Extraction (RE) is a core task in Information Extraction that identifies relationships between entities in the text. Existing prompt construction methods using Large Language Models (LLMs) achieve success by providing demonstrations to enhance relation extraction abilities. However, these methods provide retrieved demonstrations that have weak relational correlations with the test input and overlook the importance of associated entities and label semantics in RE tasks. This results in incomplete prompt construction, leading to suboptimal alignment between LLMs and RE tasks. To address these issues, we propose a Relation-Aware Retrieval Augmented Generation model (RelAwareRAG) for relation extraction. Firstly, we design a relation-aware retrieval that can identify demonstrations more effectively aligned with the target entity relation by fine-tuning a pre-trained language model for the target RE task and generating task-specific representations. Secondly, we refine the candidate relation space through entity constraint pruning and perform label semantic projection for LLMs to optimize the prompt construction, thereby improving the adaptability of LLMs in relation extraction tasks. Extensive experiments on TACRED and TACREV show that RelAwareRAG achieves superior performance, with a 3.65% increase in F1 score on TACRED and a 3.76% improvement on TACREV.

Cognitive diagnosis (CD) aims to discover students’ mastery of knowledge concepts through response logs and it is an important task in intelligence education. CD is generally performed with the advantage of students doing exercises, which is learned by linear interacting between student proficiency and exercise difficulty. However, existing methods represent the proficiency and difficulty in view of knowledge concepts independently, which is in coarse-granularity, leading to the indiscrimination for different combinations of students and exercises under the same concept. To this end, we propose a cognitive diagnosis method with nonlinear dependence between students and exercises (CDND), in which, more fine-grained information is captured to distinguish the different combinations of students and exercises. First, the nonlinear dependence is captured with a novel attention-like mechanism by interacting between students and exercises. Moreover, this nonlinear dependence is embedded as the interactive discrimination vector. Second, the interactive discrimination is used to adjust the advantage using a second-order interaction, which can strengthen the discrimination of the advantage under different combinations of students and exercises. Extensive experimental results on three real data sets validate our interactive discrimination has good compatibility with existing methods, and with it, our CDND achieves an obvious improvement in performance. Our code is available onhttps://github.com/joyce99/LinZhihao/tree/main/CDND-master.

Knowledge Tracing (KT) is a fundamental task in educational data mining that mainly focuses on tracing students’ dynamic knowledge states of skills. Bayesian Knowledge Tracing (BKT) has been widely researched and applied due to its good interpretability, using the hidden Markov model to model students’ question–answering process. Standard BKT considers only one skill in each question. To address this limitation, we proposed a Multi-skills Bayesian Knowledge Tracing (MBKT) method based on evolutionary algorithms in our previous work. MBKT employs evolutionary algorithms as the optimization method for BKT, enabling it to trace changes in students’ mastery of multiple skills simultaneously. However, MBKT has the drawback of taking too long for a single individual evaluation, and a large number of valueless individuals invoke the real evaluation process, especially when dealing with a large amount of data to be evaluated. This hinders its application in real online education scenarios. Therefore, the Surrogate Model-assisted Multi-skills Bayesian Knowledge Tracing (SA-MBKT) method is proposed to address these issues by introducing a window strategy and a surrogate model method. Extensive experiments on real-world datasets demonstrate that SA-MBKT significantly enhances temporal performance without affecting the predictive performance of the model.

Integrating multimodal data from diverse sources is crucial for enhancing various applications. Multimodal entity alignment (MMEA), which discovers equivalent entities across different sources and modalities, aims to eliminate data silos for comprehensive integration. A key challenge in MMEA is effectively fusing vector representations from different modalities of the same entity for optimal entity matching. Existing fusion methods involve individual fusion operators (e.g., concatenation and summation) or the manual design of complex network structures, incurring significant human resource costs. In this paper, for the first time, we introduce the research question of automatic fusion for MMEA and propose an efficient approach from the perspective of automated architecture search. Experimental comparisons with state-of-the-art methods on real-world datasets demonstrate the effectiveness of the proposed approach.

Tables, as an important means of data storage, are widely used in spreadsheets, web tables, and PDFs. By integrating information from table data with knowledge re-trieved from an external knowledge base, and examining the correspondences between cell values in the table and instances in the knowledge base, we can extract knowledge from the table to augment and enrich the knowledge base. To achieve this goal, we first need to classify table cells based on their functions in the layout. Due to the diverse structures arising from the arrangements of rows and columns, as well as the complexity of content resulting from concise data storage, current automation techniques heavily rely on stylistic features of table cells, such as font or color. Moreover, these methods are rarely experimented with or validated on tables without style features. Recent literature indicates that large language models (LLMs) demonstrate an ability to understand the structure and content of tables in tasks such as table judgment reasoning. Even without extensive feature inputs or pre-training, LLMs still show comparable results to machine learning and deep learning in these tasks. Therefore, this paper attempts to apply LLMs to table cell classification without using other stylistic features. We have designed a 4-component prompt paradigm (Classification Definition, Instruction, Table, Com-pletion), representing respectively the classification definition, task instructions, table data, and result output. We conduct experiments on three datasets CIUS, SAUS, and DEEX for table cell classification with one-shot learning. Our experimental results show that with the assistance of LLMs, better results can be achieved without utilizing stylistic features.

Summary In person‐centric applications, the prevalence of shared names significantly hampers document retrieval, web search, and database integration, highlighting the critical need for name disambiguation. Network embedding based unsupervised name disambiguation methods have received much attention due to their wide applicability and superior disambiguation performance. However, existing methods require complex feature engineering, such as extracting biographical and source content features or constructing supplementary features from external knowledge bases, which are unavailable in privacy‐preserving scenarios. Moreover, they may face challenges such as imbalanced node training or overlooking global statistical information during node embedding learning. In this article, we propose a method to tackle the name disambiguation problem based on high‐degree penalty and global statistical information using only relational data. First, we construct a weighted source similarity network based on multi‐hop collaborator relationships. Second, we employ a high‐degree penalty based global statistical network embedding model to learn low‐dimensional node embeddings and preserve the structural features of the network. Finally, we cluster the same sources using a joint clustering algorithm that does not require prior knowledge of the number of clusters. The experiment validates the effectiveness of the proposed method on two real‐life datasets.

The superior performance of large-scale pre-trained models, such as Bidirectional Encoder Representations from Transformers (BERT) and Generative Pre-trained Transformer (GPT), has received increasing attention in both academic and industrial research and has become one of the current research hotspots. A pre-trained model refers to a model trained on large-scale unlabeled data, whose purpose is to learn general language representation or features for fine-tuning or transfer learning in subsequent tasks. After pre-training is complete, a small amount of labeled data can be used to fine-tune the model for a specific task or domain. This two-stage method of “pre-training+fine-tuning” has achieved advanced results in natural language processing (NLP) tasks. Despite widespread adoption, existing fixed fine-tuning schemes that adapt well to one NLP task may perform inconsistently on other NLP tasks given that different tasks have different latent semantic structures. In this article, we explore the effectiveness of automatic fine-tuning pattern search for layer-wise learning rates from an evolutionary optimization perspective. Our goal is to use evolutionary algorithms to search for better task-dependent fine-tuning patterns for specific NLP tasks than typical fixed fine-tuning patterns. Experimental results on two real-world language benchmarks and three advanced pre-training language models show the effectiveness and generality of the proposed framework.

In educational data mining, knowledge tracing (KT) aims to model learning performance based on student knowledge mastery. Deep-learning-based KT models perform remarkably better than traditional KT and have attracted considerable attention. However, most of them lack interpretability, making it challenging to explain why the model performed well in the prediction. In this paper, we propose an interpretable deep KT model, referred to as fuzzy deep knowledge tracing (FDKT) via fuzzy reasoning. Specifically, we formalize continuous scores into several fuzzy scores using the fuzzification module. Then, we input the fuzzy scores into the fuzzy reasoning module (FRM). FRM is designed to deduce the current cognitive ability, based on which the future performance was predicted. FDKT greatly enhanced the intrinsic interpretability of deep-learning-based KT through the interpretation of the deduction of student cognition. Furthermore, it broadened the application of KT to continuous scores. Improved performance with regard to both the advantages of FDKT was demonstrated through comparisons with the state-of-the-art models.

Cognitive Diagnostic Models (CDMs) aim to analyze students' cognitive levels of each knowledge component (KC) by mining educational data. Existing CDMs can be mainly divided into two categories, i.e., traditional probability-based and neural-network-based. Most probabilistic models have the advantages of simplicity and good interpretability, but suffer from slow training time in the case of a large number of KCs. Neural-network-based methods are widely considered to be superior to probabilistic models due to their good performance. However, neural network methods are less interpretable than probabilistic models, thus limiting their usefulness in practice. Because most existing probabilistic models are optimized iteratively based on single-point-based search methods, they may be easily trapped in local optimum due to the influence of the initial points. And evolutionary algorithms (EAs) have good global search ability. Therefore, an interesting question is whether a simple probabilistic model based on evolutionary optimization can rival neural-network in limited optimization time. Thus, a hybrid EA with a customized local search is proposed. Experimental results on three real-world datasets show that our method outperforms the compared 7 models (including 2 state-of-the-art neural-network-based models); and the running time of our method is significantly less than the compared probabilistic models.

Point-of-Interest (POI) recommendation, pivotal for guiding users to their next interested locale, grapples with the persistent challenge of data sparsity. Whereas knowledge graphs (KGs) have emerged as a favored tool to mitigate the issue, existing KG-based methods tend to overlook two crucial elements: the intention steering users’ location choices and the high-order topological structure within the KG. In this paper, we craft an Intention-aware Knowledge Graph (IKG) that harmonizes users’ visit histories, movement trajectories, and location categories to model user intentions. Building upon IKG, our novel Intention-aware Knowledge Graph Network (IKGN) delves deeper into the POI recommendation by weighing and propagating node embeddings through an attention mechanism, capturing the unique locational intent of each user. A sequential model like GRU is then employed to ensure a comprehensive representation of users’ short- and long-term location preferences. An empirical study on two real-world datasets validates the effectiveness of our proposed IKGN, with it markedly outshining seven benchmark rival models in both Recall and NDCG metrics. The code of IKGN is available at https://github.com/Jungle123456/IKGN.

Event detection (ED) is a key subtask of information extraction to extract key events, such as stock rise and fall and social public opinion, from news or social media. Although current GCN-based event detection methods achieve remarkable success via building graphs with dependency trees, they typically suffer from two challenges: 1) They use sequence models to learn contextual information of sentences, ignoring the longterm dependencies problem of sequence models might learn ineffective information and make it propagate in GCN layers. 2) Most methods do not exploit global dependency label information and grammatical structure information that convey rich linguistic knowledge directly, and only consider local dependency label information. To cope with these challenges, we propose a novel event detection model via semantic-reconstructed graph transformer networks (SRGTNED), which incorporates semantic reconstruction and path information collection methods. Using the semantic reconstruction method, we assign a pruned sequence to each word based on the path information to capture contextual information consistent with sentence semantics. Moreover, to better utilize global dependency label information and grammatical structure information, a Graph Transformer Network (GTN)-based heterogeneous graph embedding framework is introduced to automatically learn path information between important words by converting sentences as heterogeneous graphs. We conduct experiments on the ACE2005 dataset and the Commodity News dataset, and the experimental results demonstrate that our method significantly outperforms 11 state-of-the-art baselines in terms of the F1-score.

Multivariate time series classification aims to determine the labels for multivariate time series samples. Although variable interaction relationships and sample similarity relationships exist in multivariate time series, the available related methods usually ignore the rich relationships and are ineffective in exploiting these. To solve this problem, we propose a Hierarchical Graph Embedding for Multivariate Time Series Classification (MTSC-HGE), which consists of a variable-wise attentive graph pooling module and a sample-wise graph convolutional module to obtain the relationships of variables and samples. Specifically, we design an attentive graph pooling module based on self-attention, which can obtain sample features fusing temporal patterns and variable interaction relationships in samples. Furthermore, we propose a graph mapping criterion that converts the MTS dataset into a graph based on dynamic time warping to explicitly reflect the similarity relationships between samples. To capture latent sample relationships, a GCN module is utilized on the sample graph to integrate sample features obtained from the attentive graph pool module. In addition, a classifier takes the rich representation output by the model to get the final predicted class. Extensive experiments on 14 public datasets show that MTSC-HGE significantly outperforms state-of-the-art baselines.

Exercise recommendation is a fundamental and important task in the E-learning system, facilitating students' personalized learning. Most existing exercise recommendation algorithms design a scoring criterion (e.g., weakest mastery, lowest historical correctness) in conjunction with experience, and then recommend the recommended knowledge concepts (KCs). These algorithms rely entirely on the scoring criteria by treating exercise recommendations as a centralized system. However, it is a complex problem for the centralized system to choose a limited number of exercises in a period of time to consolidate and learn the KCs efficiently. Moreover, different groups of students (e.g., different countries, schools, or classes) have different solutions for the same group of KCs according to their own situations, in the spirit of competency-based instructing. Therefore, we propose Meta Multi-Agent Exercise Recommendation (MMER). Specifically, we design the multi-agent exercise recommendation module, in which the KCs involved in exercises are considered agents with competition and cooperation among them. And the meta-training stage is designed to learn a robust recommendation module for new student groups. Extensive experiments on real-world datasets validate the satisfactory performance of the proposed model. Furthermore, the effectiveness of the multi-agent and meta-training part is demonstrated for the model in recommendation applications.

Large-scale pre-trained models such as GPT and BERT have demonstrated remarkable performance in information extraction tasks. However, their black-box nature poses challenges for reliability and interpretability. In contrast, rule- based extraction methods have better interpretability, but typically require domain experts to manually establish rules, limiting their generalization ability. In industry, there is often a demand for reliable knowledge extraction to reduce the time spent on manual verification of each piece of knowledge. In this paper, we explore the idea of combining GPT and symbolic-based methods to automatically discover reliable extraction patterns in text with a particular writing style. This method leverages the characteristics of high information density and similar writing patterns in text with a specific writing style to generate verifiable and reliable patterns. We conduct experiments on two datasets with a specific writing style to demonstrate its effectiveness, validating the idea of combining large models for reliable information extraction pattern discovery in the tested datasets.

Nature-inspired population-based stochastic search algorithms (SSA) have demonstrated effectiveness in solving many real-world dynamic optimization problems (DOPs), such as dynamic optimal power flow (DOPF) problems. The basic idea of solving DOPs using SSAs is to “track the moving optima”, rather than solving the changed problems from scratch. Its hidden assumption is that the problems are slightly changed in general, and it is expected that the search process can be accelerated by learning from past search experiences. However, the hidden assumption that the current search process can always benefit from previous search experiences might not be true for certain cases. For example, if the current problem is not similar to any of its previous problems, the historical solutions might not be helpful for the current search. To reduce such negative transfer, the following issues are worthy of study: how to choose which historical problems to learn from, and how to determine the degree of learning from historical problems. To solve the above problems, we propose a dynamic processing framework based on fuzzy rules from the perspective of incorporating human knowledge for optima tracking. The experimental results on the DOPF problems show that the SSA with the proposed optima tracking strategy outperforms other comparative algorithms. We open up the code and data of our algorithm44https://github.com/DMiC-Lab-HFUT/SMDE-Transfer.

Genealogical knowledge graphs depict the relationships of family networks and the development of family histories. They can help researchers to analyze and understand genealogical data, search for genealogical descendant paths, and explore the origins of a family more easily. However, the heterogenous, autonomous, complex, and evolving natures of genealogical data bring challenges to the development of contemporary genealogical knowledge graph models. Applying existing methods to genealogical data may be improper because general knowledge graph models lack in-depth domain knowledge. In this paper, we propose a genealogical knowledge graph model named Huapu-KG that combines HAO intelligence (human intelligence + artificial intelligence + organizational intelligence) to implement the construction and applications of genealogical knowledge graphs. Furthermore, challenges in constructing genealogical knowledge graphs are demonstrated, and experiments conducted on real-world genealogical datasets verify the feasibility and effectiveness of our proposed model.

Investment promotion refers to the process by which the government uses disposable resources to attract investors to the region for production and business activities. The existing basic mode of attracting investment is to collect information about enterprises and entrepreneurs through manual methods, determine the target enterprise from the list of enterprises, and then attract investment through visits, negotiations, and other methods. As contacting and visiting companies one by one requires huge amounts of manpower and time, the choice of target companies is critical for attracting investments. However, to the best of our knowledge, no study has conducted research from the perspective of knowledge-graph-based recommendation. In this study, we define the problem of target company recommendation in the process of investment promotion, and analyze the characteristics of the problem and the challenges it faces based on the background of the actual problem. Then, a two-tier model for solving this problem is provided from the perspective of knowledge graph reasoning. Aiming at the problem that the knowledge graph will frequently change, the model is designed based on the idea of combining the advantages of global and local link prediction. The experimental results on real-world data demonstrate the effectiveness of the proposed model.

In the era of big data, fragmented knowledge, multisource heterogeneity, and different representation forms of the same entities in various data sources have posed considerable challenges to entity fusion. How to effectively integrate multisource knowledge for the same entities has provoked vast amounts of attention and research from multiple disciplines. Most existing methods for entity fusion can be categorized into two classes: one is to establish an association between the same entities, and the other is to delete duplicate entities after knowledge fusion and create a new fusion entity. However, in these two classes of methods, the former does not achieve true knowledge fusion and semantic interoperability, while the latter may cause irreversible loss of original information. In this paper, we propose a novel entity fusion scheme: Hypernode. Hypernode fuses the same entity in different data sources into a new entity while retaining the original data. We verify the effectiveness of Hypernode on multiple models of link prediction experiments. Several practical application cases illustrate the applicability of Hypernode in data traceability, open domain knowledge fusion, and multi-modal knowledge graph fusion.

Zero-shot learning aims to recognize unseen-classes using some seen-class samples as training set. It is challenging owing to that the feature representations of unseen-class samples are unavailable. Existing methods transfer the mapping from seen-classes to unseen-classes with the correlation as a bridge, in which, the semantic representations are used to discriminate the classes. However, the unavailability of visual representations for unseen-classes and the insufficient discrimination of semantic representations make the zero-shot learning challenging. Therefore, the visual representations are learned as complements to semantic representations to construct a multi-modal knowledge graph (KG), and a zero-shot learning method based on multi-modal KG is proposed in this paper. Specially, a semantic KG is introduced to capture the correlation of classes, and with the correlation, the visual feature representations of all classes are learned. Then, the discriminative visual representations and the semantic representations are used together to construct a multi-modal KG. With the multi-modal KG, the classifier for seen-classes is transferred to unseen classes. Extensive experimental results show the effectiveness of our method.

Gradient-based repair aims to repair infeasible solutions to feasible ones using the gradient information of the constraints. As an effective constraint handling method, gradientbased repair has received extensive attention and has been applied in various evolutionary algorithms (EAs). Nevertheless, due to the complexity of constraints in practical problems, a single infeasible solution often needs to be repaired multiple times until it becomes a feasible solution or reaches the maximum number of repairs. As far as we know, existing related research on gradient-based repair mainly applies this method directly to EAs, while there is little work in the evolutionary computing community on how to improve gradient-based repair. Currently, the multiple repairs for a single individual are independent. That is, the current repair does not consider the previous repair experience. However, only using gradient information to repair infeasible individuals may result in oscillations in the search process. Therefore, in this paper, we propose a heuristic gradient-based repair method (HGR) which exploits the previous repair information of an individual to alleviate this issue. Experimental results on several benchmarks demonstrate the effectiveness of the proposed method. The source code is available at https://github.com/DMiC-Lab-HFUT/HGR-SMC2022.

Online education promotes the sharing of learning resources. Knowledge tracing (KT) is aimed at tracking the cognition function of students according to their performance on various exercises at different times and has attracted considerable attention. Existing KT models primarily use bisection representations for the performance and cognitive states of students, thus limiting the application scope of these models and the accuracy of the evaluation of student cognitive performance in learning processes. Therefore, fuzzy Bayesian KT models (namely, FBKT and T2FBKT) are proposed to address continuous score scenarios (e.g., subjective examinations) so that the applicability of KT models may be broadened. Moreover, fine-grained cognitive states can be discerned. In particular, referring to type-2 fuzzy theory, T2FBKT mitigates the model uncertainty of FBKT induced by uncertain parameters. Finally, extensive experiments demonstrate the effectiveness of the proposed fuzzy KT models.

The automatic construction of knowledge graphs (KGs) from multiple data sources has received increasing attention. The automatic construction process inevitably brings considerable noise, especially in the construction of KGs from unstructured text. The noise in a KG can be divided into two categories: factual noise and low-quality noise. Factual noise refers to plausible triples that meet the requirements of ontology constraints. For example, the plausible triple satisfies the constraints that the head entity “New_York” is a city and the tail entity “America” belongs to a country. Low-quality noise denotes the obvious errors commonly created in information extraction processes. This study focuses on entity type errors. Most existing approaches concentrate on refining an existing KG, assuming that the type information of most entities or the ontology information in the KG is known in advance. However, such methods may not be suitable at the start of a KG's construction. Therefore, the authors propose an effective framework to eliminate entity type errors. The experimental results demonstrate the effectiveness of the proposed method.

Graph edge partitioning, which is essential for the efficiency of distributed graph computation systems, divides a graph into several balanced partitions within a given size to minimize the number of vertices to be cut. Existing graph partitioning models can be classified into two categories: offline and streaming graph partitioning models. The former requires global graph information during the partitioning, which is expensive in terms of time and memory for large-scale graphs. The latter creates partitions based solely on the received graph information. However, the streaming model may result in a lower partitioning quality compared with the offline model. Therefore, this study introduces a Local Graph Edge Partitioning model, which considers only the local information (i.e., a portion of a graph instead of the entire graph) during the partitioning. Considering only the local graph information is meaningful because acquiring complete information for large-scale graphs is expensive. Based on the Local Graph Edge Partitioning model, two local graph edge partitioning algorithms—Two-stage Local Partitioning and Adaptive Local Partitioning—are given. Experimental results obtained on 14 real-world graphs demonstrate that the proposed algorithms outperform rival algorithms in most tested cases. Furthermore, the proposed algorithms are proven to significantly improve the efficiency of the real graph computation system GraphX.

Considering the uncertainty of power generations, in addition to the uncertainty of loads, is more and more important because of the increasing use of renewable energy sources. Most existing works on dynamic optimal power flow (DOPF) have only focused on either the uncertainty of loads (called demand-side uncertainty) or the uncertainty of power generations (called supply-side uncertainty). As far as we know, only a little work on the dynamic OPF problems considered both uncertainties simultaneously. It might be because the combination of variable uncertainties could lead to a huge problem size for existing methods. In this paper, inspired by the ideas in the field of evolutionary dynamic optimization (EDO), we attempt to deal with uncertain parameters from the perspective of tracking the moving optimum. A species and memory enhanced differential evolutionary algorithm (called SMDE) is specially designed to solve the DOPF with double-sided uncertainties. Specifically, in order to deal with the double-sided uncertainties, a modified memory strategy and an improved multi-population strategy were introduced, where the multi-population strategy includes two versions. The experimental results on the modified IEEE 57-bus and 118-bus systems show that the proposed algorithms perform much better than the comparison algorithms for most cases.

Graph edge partitioning divides the edges of an input graph into multiple balanced partitions of a given size to minimize the sum of vertices that are cut, which is critical to the performance of distributed graph computation platforms. Existing graph partitioning methods can be classified into two categories: offline graph partitioning and streaming graph partitioning. The first category requires global information for a graph during the partitioning, which is expensive in terms of time and memory for large-scale graphs. The second category, however, creates partitions solely based on the received edge information, which may result in lower performance than the offline methods. Therefore, in this study, the concept of local graph partitioning is introduced from local community detection to consider only local information, i.e., a part of the graph, instead of the graph as a whole, during the partitioning. The characteristic of storing only local information is important because real-world graphs are often large in scale, or they increase incrementally. Based on this idea, we propose a two-stage local partitioning algorithm, where the partitioning process is divided into two stages according to the structural changes of the current partition, and two different strategies are introduced to deal with the respective stages. Experimental results with real-world graphs demonstrate that the proposed algorithm outperforms the rival algorithms in most cases, including the state-of-the-art algorithm METIS.

Dynamic constrained optimization problems (DCOPs) are difficult to solve because both objective function and constraints can vary with time. Although DCOPs have drawn attention in recent years, little work has been performed to solve DCOPs with multiple dynamic feasible regions from the perspective of locating and tracking multiple feasible regions in parallel. Moreover, few benchmarks have been proposed to simulate the dynamics of multiple disconnected feasible regions. In this paper, first, the idea of tracking multiple feasible regions, originally proposed by Nguyen and Yao, is enhanced by specifically adopting multiple subpopulations. To this end, the dynamic species-based particle swam optimization (DSPSO), a representative multipopulation algorithm, is adopted. Second, an ensemble of locating and tracking feasible regions strategies is proposed to handle different types of dynamics in constraints. Third, two benchmarks are designed to simulate the DCOPs with dynamic constraints. The first benchmark, including two variants of G24 (called G24v and G24w), could control the size of feasible regions. The second benchmark, named moving feasible regions benchmark (MFRB), is highly configurable. The global optimum of MFRB is calculated mathematically for experimental comparisons. Experimental results on G24, G24v, G24w, and MFRB show that the DSPSO with the ensemble of strategies performs significantly better than the original DSPSO and other typical algorithms.

Clustering is a classical unsupervised learning task, which is aimed to divide a data set into several groups with similar objects. Clustering problem has been studied for many years, and many excellent clustering algorithms have been proposed. In this paper, we propose a novel clustering method based on density, which is simple but effective. The primary idea of the proposed method is given as follows. Firstly, the point with the largest local density in a cluster is considered as the cluster center. The local density of each point is estimated based on the distance (called radius) between the point and its k-th nearest neighbor. The point with a smaller radius indicates a larger local density. Secondly, the difference of the local densities between each two internal points should be small, while the difference between the density of a border point and the density of an internal point should be relatively large. Thirdly, if the intersection of k nearest neighbors of two points is small, they should be assigned to different clusters. The proposed algorithm has been compared with a typical clustering algorithm named FDPCluster, and the experimental results show that our algorithm has better clustering quality.

Evolutionary Algorithms (EAs) with gradient-based repair, which utilize the gradient information of the constraints set, have been proved to be effective. It is known that it would be time-consuming if all infeasible individuals are repaired. Therefore, so far the infeasible individuals to be repaired are randomly selected from the population and the strategy of choosing individuals to be repaired has not been studied yet. In this paper, the Species-based Repair Strategy (SRS) is proposed to select representative infeasible individuals instead of the random selection for gradient-based repair. The proposed SRS strategy has been applied to εDEag which repairs the random selected individuals using the gradient-based repair. The new algorithm is named SRS-εDEag. Experimental results show that SRS-εDEag outperforms εDEag in most benchmarks. Meanwhile, the number of repaired individuals is reduced markedly.