As an emerging weakly supervised learning framework, partial label learning aims to induce a multi-class classifier from ambiguous supervision information where each training example is associated with a set of candid...
详细信息
As an emerging weakly supervised learning framework, partial label learning aims to induce a multi-class classifier from ambiguous supervision information where each training example is associated with a set of candidate labels, among which only one is the true label. Traditional feature selection methods, either for single label and multiple label problems, are not applicable to partial label learning as the ambiguous information contained in the label space obfuscates the importance of features and misleads the selection process. This makes the selection of a proper feature subset from partial label examples particularly challenging, and therefore has rarely been investigated. In this paper, we propose a novel feature selection algorithm for partial label learning, named PLFS, which considers not only the relationships between features and labels, but also exploits the relationships between instances to select the most informative and important features to enhance the performance of partial label learning. PLFS constructs an adaptive weighted graph to exploit the similarity information among instances, differentiate the label space and weight the feature space, which leads to the selection of a proper feature subset. Extensive experiments over a broad range of benchmark data sets clearly validate the effectiveness of our proposed feature selection approach.
Kolmogorov-Arnold Networks (KAN) is an emerging neural network architecture in machine learning. It has greatly interested the research community about whether KAN can be a promising alternative to the commonly used M...
详细信息
Analyzing short texts infers discriminative and coherent latent topics that is a critical and fundamental task since many real-world applications require semantic understanding of short texts. Traditional long text to...
详细信息
Conversational emotion recognition (CER) is an important research topic in human-computer interactions. Although recent advancements in transformer-based cross-modal fusion methods have shown promise in CER tasks, the...
详细信息
Few-shot knowledge graph completion (FKGC) aims to infer unknown fact triples of a relation using its few-shot reference entity pairs. Recent FKGC studies focus on learning semantic representations of entity pairs by ...
详细信息
Identifying semantic types for attributes in relations,known as attribute semantic type(AST)identification,plays an important role in many data analysis tasks,such as data cleaning,schema matching,and keyword search i...
详细信息
Identifying semantic types for attributes in relations,known as attribute semantic type(AST)identification,plays an important role in many data analysis tasks,such as data cleaning,schema matching,and keyword search in ***,due to a lack of unified naming standards across prevalent information systems(*** islands),AST identification still remains as an open *** tackle this problem,we propose a context-aware method to figure out the ASTs for relations in this *** transform the AST identification into a multi-class classification problem and propose a schema context aware(SCA)model to learn the representation from a collection of relations associated with attribute values and schema *** on the learned representation,we predict the AST for a given attribute from an underlying relation,wherein the predicted AST is mapped to one of the labeled *** improve the performance for AST identification,especially for the case that the predicted semantic types of attributes are not included in the labeled ASTs,we then introduce knowledge base embeddings(***)to enhance the above representation and construct a schema context aware model with knowledge base enhanced(SCA-KB)to get a stable and robust *** experiments based on real datasets demonstrate that our context-aware method outperforms the state-of-the-art approaches by a large margin,up to 6.14%and 25.17%in terms of macro average F1 score,and up to 0.28%and 9.56%in terms of weighted F1 score over high-quality and low-quality datasets respectively.
Blockchain technologies pave a promising way for implementing the inter-organizational processes. Most of the current research works translate the execution logic in the process models into the smart contracts, which ...
详细信息
Multi-modal entity alignment (MMEA) aims to identify equivalent entities across different multi-modal knowledge graphs. In these graphs, entities are enriched with information from various modalities, such as text, im...
详细信息
ISBN:
(数字)9798331508821
ISBN:
(纸本)9798331508838
Multi-modal entity alignment (MMEA) aims to identify equivalent entities across different multi-modal knowledge graphs. In these graphs, entities are enriched with information from various modalities, such as text, images, and numerical data, making the alignment task both challenging and crucial for improving knowledge graph quality. Current MMEA algorithms typically encode entity information separately for each modality using corresponding encoders, and then integrate these representations through various modal fusion strategies. However, these methods often fail to fully exploit the multi-modal information of entities. To address this issue, we propose a feature-enhanced multi-modal entity alignment transformer (FEMEAT). FEMEAT enhances entity attribute information by incorporating modal distribution data, which captures the inherent distribution of different modalities for each entity. This inclusion allows the model to have a richer understanding of entity characteristics across modalities. Additionally, FEMEAT utilizes an Optical Character Recognition (OCR) model to extract and incorporate textual information from images. By integrating this text extracted from images, the model can better utilize the visual modality, enhancing its ability to understand and process multi-modal information. Furthermore, FEMEAT employs a multi-head cross-modal attention (MHCA) mechanism for modal fusion to achieve comprehensive multi-modal entity representation. This mechanism enables the model to attend to different modalities simultaneously and learn a detailed representation of entities by considering the interactions between modalities. The multi-head cross-modal attention mechanism facilitates a nuanced understanding and integration of multi-modal data. Experimental results demonstrate that our model achieves state-of-the-art (SOTA) performance across various training scenarios. The code and datasets used in this study can be accessed at https://***/zewenD/FEMEAT.
Graph pattern matching is a technique widely used in various fields such as protein structure analysis, social group querying, and expert localization. This technique involves finding matching subgraphs in large socia...
详细信息
ISBN:
(数字)9798350373554
ISBN:
(纸本)9798350373561
Graph pattern matching is a technique widely used in various fields such as protein structure analysis, social group querying, and expert localization. This technique involves finding matching subgraphs in large social networks that align with the patterns specified in the pattern graph. In this paper, we focus on a specific sub-problem in social group querying, known as the cooperative team query, which arises from practical applications, where the nodes in the pattern graph and the data graph represent team member entities, while the edges represent their social relationships. We note that the requirements of many teams in the real world are dynamic, necessitating iterative computation for graph pattern matching using traditional methods. To address this challenge in highly dynamic systems, we propose a graph pattern matching method based on core pattern graph matching cache. This approach involves extracting the core pattern graph, and comprising core team members based on the characteristics of cooperative teams. The core graph-based matching cache enables the second half of the algorithm to operate on an order-of-magnitude smaller graph, significantly improving efficiency. Additionally, the multi-threaded approach fully leverages hardware resources, synchronizing multiple matching result of the core pattern graph to reduce matching time. Experimental results on three real social network datasets demonstrate that our proposed algorithm, Core Pattern Graph Matching Cache-based Multi-threaded Exploration (CCMTE), significantly outperforms existing methods in terms of efficiency.
The recent rise of conversational applications such as online customer service systems and intelligent personal assistants has promoted the development of conversational knowledge base question answering (ConvKBQA). D...
详细信息
暂无评论