Low-rank tensor approximation with local structure for multi-view intrinsic subspace clustering

Fu, Lele; Yang, Jing‐Hua; Chen, Chuan; Zhang, Chuanfu

doi:10.1016/j.ins.2022.05.091

Cited by 31 publications

(4 citation statements)

References 67 publications

(48 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…F-score=2 × (precision × recall)/(precision+recall ) (6) In information retrieval, precision is defined as purity of the retrieval and Recall identifies the percentage of completeness of retrieval. There is opposite relationship between precision and recall where one increases other decreases.…”

Section: Evaluation Metricsmentioning

confidence: 99%

“…During keyphrase, extraction operation content of the web page is divided into tokens or words after undergoing many sub-steps different weights (0-1) are received from each token or word. The word with high weight considered being more relevant to web content or subject matter and lower weight consider being less association of the word with web content [5][6][7]. For example, the sports webpage of cricket games includes keyphrases "umpire" and "car" where the umpire is relevant, and the car is not relevant to the content of the webpage.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

W-rank: A keyphrase extraction method for webpage based on linguistics and DOM-base features

Shah,

Ahmed,

Sathio

et al. 2023

VAWKUM trans. comput. sci.

View full text Add to dashboard Cite

This paper addresses the problem of an automatic keyphrase extraction for a webpage text. Our method is unsupervised, and we call it W-rank. In our method, first we extract the text of a webpage and tokenize into three different candidate words list: unigram ,bigrams and noun phrases. Then we assign score to all words based on their individual appearance in linguistic and DOM-based feature sets. In the final step, we rank these candidate words using score and select top 5 keyphrase from each list and combine them as a final keyphrases for a given webpage. We focus more on the relevancy of keyphrases to its content using linguistic features. We compare our method with other methods using precision, recall and f-score. The experimental result shows, W-rank improves the performance of our previous method D-rank and outperforms other state of art methods.

show abstract

Section: Evaluation Metricsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

W-rank: A keyphrase extraction method for webpage based on linguistics and DOM-base features

Shah,

Ahmed,

Sathio

et al. 2023

VAWKUM trans. comput. sci.

View full text Add to dashboard Cite

show abstract

“…Several methods take advantage of this property. Fu et al [5] develop a low-rank tensor approximation model for multiview intrinsic subspace clustering that effectively reduces view-specific constraints and improves optimization, with notable success on real-world datasets. Wang et al's [6] tensor low-rank and sparse representation method skillfully preserves intrinsic 3D structures in hyperspectral anomaly detection.…”

Section: Introduction 1background and Related Workmentioning

confidence: 99%

Color Image Recovery Using Generalized Matrix Completion over Higher-Order Finite Dimensional Algebra

Liao,

Guo,

Gao

et al. 2023

Axioms

View full text Add to dashboard Cite

To improve the accuracy of color image completion with missing entries, we present a recovery method based on generalized higher-order scalars. We extend the traditional second-order matrix model to a more comprehensive higher-order matrix equivalent, called the “t-matrix” model, which incorporates a pixel neighborhood expansion strategy to characterize the local pixel constraints. This “t-matrix” model is then used to extend some commonly used matrix and tensor completion algorithms to their higher-order versions. We perform extensive experiments on various algorithms using simulated data and publicly available images. The results show that our generalized matrix completion model and the corresponding algorithm compare favorably with their lower-order tensor and conventional matrix counterparts.

show abstract

“…They construct the self-representation matrices for all views and perform the single-view clustering method, e.g., k-means, on the consensus representation matrix for the cluster partition. Moreover, tensor learning-based multi-view clustering methods [32], [33], [34], [35] also show promising clustering performance on multiview graph data, which stack the representation matrices into a three-order tensor to explore the high-order correlation among all views. Despite great success, they only consider the topological structures, i.e, the adjacent matrices of the graph data, and pay little attention to the node attribute information [36].…”

Section: Introductionmentioning

confidence: 99%

Multi-view Attributed Graph Clustering

Lin

Kang

Zhang

et al. 2021

IEEE Trans. Knowl. Data Eng.

View full text Add to dashboard Cite

Multi-view attributed graph clustering is an important approach to partition multi-view data based on the attribute feature and adjacent matrices from different views. Some attempts have been made in utilizing Graph Neural Network (GNN), which have achieved promising clustering performance. Despite this, few of them pay attention to the inherent specific information embedded in multiple views. Meanwhile, they are incapable of recovering the latent high-level representation from the low-level ones, greatly limiting the downstream clustering performance. To fill these gaps, a novel Dual Information enhanced multi-view Attributed Graph Clustering (DIAGC) method is proposed in this paper. Specifically, the proposed method introduces the Specific Information Reconstruction (SIR) module to disentangle the explorations of the consensus and specific information from multiple views, which enables GCN to capture the more essential low-level representations. Besides, the Mutual Information Maximization (MIM) module maximizes the agreement between the latent high-level representation and lowlevel ones, and enables the high-level representation to satisfy the desired clustering structure with the help of the Self-supervised Clustering (SC) module. Extensive experiments on several realworld benchmarks demonstrate the effectiveness of the proposed DIAGC method compared with the state-of-the-art baselines.

show abstract

Low-rank tensor approximation with local structure for multi-view intrinsic subspace clustering

Cited by 31 publications

References 67 publications

W-rank: A keyphrase extraction method for webpage based on linguistics and DOM-base features

W-rank: A keyphrase extraction method for webpage based on linguistics and DOM-base features

Color Image Recovery Using Generalized Matrix Completion over Higher-Order Finite Dimensional Algebra

Multi-view Attributed Graph Clustering

Contact Info

Product

Resources

About