Document clustering using nonnegative matrix factorization

By Farial Shahnaz, Michael W. Berry, V. P. Pauca, Robert J. Plemmons. Information Processing & Management, 2006

681 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/a74534569c5532ea62da572e3b617dedc825b262

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Jiawei Han
2. Deng Cai
3. Philip S. Yu
4. Karthik Devarajan
5. Haesun Park
6. Xuelong Li
7. Xiaofei He
8. Jingu Kim

In Computer Vision and Pattern Recognition

1. Dacheng Tao
2. Lei Zhang
3. Xuelong Li
4. Deng Cai
5. Thomas S. Huang
6. Luc Van Gool
7. Haesun Park
8. Shuicheng Yan

Citations in the corpus

• Algorithms and applications for approximate nonnegative matrix factorization (2007, methodology): “However, as discussed in Shahnaz et al. (2006) for standard topic detection benchmark collections (with human-curated document clusters), the accuracy of the CNMF algorithm for document clustering degrades as the rank k increases or if the sizes of the clusters become greatly imbalanced.”
• Generalized Low Rank Models (2014, background): “However, these problems have a rich analytical structure which can sometimes be exploited [49, 10, 31], and a wide range of uses in practice [85, 126, 7, 151, 77, 47].”
• Non-negative Matrix Factorization on Manifold (2008, background): “Previous studies show that NMF is very powerful on clustering [29, 24].”
• Nonnegative Matrix Factorization: An Analytical and Interpretive Tool in Computational Biology (2008, methodology): “Gaoand Church applied the method outlined in [20] to cancer microarray data and explicitly enforced sparseness via the sum of squares of the entries of H .”
• A Semi-NMF-PCA Unified Framework for Data Clustering (2017, background): “At present, NMF and its variants have already found a wide spectrum of applications in several areas such as pattern recognition and feature extraction [23], [25], [29], dimensionality reduction, segmentation and clustering [4], [14], [18], text mining [27] and neurobiology [7].”
• Multiple Incomplete Views Clustering via Weighted Nonnegative Matrix Factorization with L2, 1 Regularization (2015, methodology): “NMF has been successfully used in unsupervised learning [31,36].”
• Improving molecular cancer class discovery through sparse non-negative matrix factorization (2005, positive): “Specifically, we adopt the point-count regularization approach that enforces sparseness of H by penalizing the number of non-zero entries rather than the sum of entries SH ij in H (Hoyer, 2002, 2004; Shahnaz et al., 2004).”
• Graph Regularized Nonnegative Matrix Factorization for Data Representation (2011, background): “Previous studies show that NMF is very powerful on document clustering [18, 14].”
• Nonnegative matrix factorization for spectral data analysis (2006, methodology): “Work on applications to air emission quality [6] and on text mining [23, 26] will also be continued.”
• K-SVD and its non-negative variant for dictionary design (2005, background): “Application for NMF are many and include dimensionality reduction [17] and analysis of data such as audio [18], text [19] and even data obtained from astronomical spectrometers [20].”
• SVD based initialization: A head start for nonnegative matrix factorization (2008, methodology): “Another important issue addressed by researchers is the incorporation of further constraints appropriate for the problem; see e.g. [6] [7] [13] [15].”
• Discovery of multi-dimensional modules by integrative analysis of cancer genomic data (2012, methodology): “Several algorithms have been developed to optimize the NMF problem (19).”
• The Why and How of Nonnegative Matrix Factorization (2014, background): “This decomposition can be interpreted as follows (see, also, e.g., [79, 101, 3]): • Because W is nonnegative, each column of W can be interpreted as a document, that is, as a bag of words.”
• Sparse Nonnegative Matrix Factorization for Clustering (2008, methodology): “[17] used NMF for text clustering and reported superior performance, and Brunet et al.”
• Constrained Nonnegative Matrix Factorization for Image Representation (2012, background): “In real applications such as image processing [2], face recognition [12], [25], and document clustering [13], [14], we usually set k ( m and k ( n.”
• Nonnegative shared subspace learning and its application to social media retrieval (2010, background): “…a data matrix X into a product of a matrix F whose columns span the latent subspace and an encoding matrix where X is a M × N nonnegative data matrix containing N documents in terms of M vocabulary words , F ( M × R nonnegative matrix) represents R basis vectors and H ( R × N nonnegative…”
• Beyond News Contents: The Role of Social Context for Fake News Detection (2017, background): “Recently, it has been shown that nonnegative matrix factorization (NMF) algorithms are very practical and popular to learn document representations [20, 28, 38].”