All dataset files contain two attributes, X and y.
Xis the multi-view data as a cell, each element in this cell is an N-by-D matrix Xk, where N is the number of data points and D is the feature dimensions. So rows of Xk correspond to data points.yis the vector of ground-truth labels.
Note: Due to Github's limitation on file size, files larger than 20MB will be stored in other repositories：
| Dataset | Instances | Clusters | Views(dimension) | Description | Type |
|---|---|---|---|---|---|
| 3Sources | 169 | 6 | Reuters(3068), BBC(3560), Guardian(3631) | A new multi-view text dataset collected from three well-known online news sources: BBC, Reuters,and The Guardian [14]. | Text |
| 100Leaves | 1600 | 100 | SD(64), FSM(64), TH(64) | Sixteen samples of leaf each of one-hundred plant species [16]. | Image |
| ACM | 3025 | 3 | View1(1870), View2(3025), View2(3025), View2(3025), View2(3025) | The dataset extracts papers published in KDD, SIGMOD, SIGCOMM, MobiCOMM, and VLDB. | |
| ALOI-1k | 110250 | 1000 | CS(77), Haralick(13), HSV(64), RGB(125) | The Amsterdam Library of Object Images is a collection of 110250 images of 1000 small objects, taken under various light conditions and rotation angles [25]. | Object |
| ALOI | 10800 | 100 | CS(77), Haralick(13), HSV(64), RGB(125) | A subset of ALOI-1k [25]. | Object |
| Animal | 11673 | 20 | View1(2688), View2(2000), View3(2001), View4(2000), | A selected set of the Animals with Attributes dataset [12]. | Animal |
| BBCSport | 544 | 5 | View1(3183), View2(3203) | This document dataset contains 544 documents from the BBC Sport website, and they are aboutthe sports news between 2004 and 2005 [4]. | Text |
| BBC4view | 685 | 5 | View1(4659), View2(4633), View3(4665), View4(4684) | Similar to BBCSport, this dataset consists of 685 documents from the BBC Sport website about sports news [4]. | Text |
| COIL20 | 1440 | 20 | Intensity(1024), LBP(944), Gabor(4096) | It is the abbreviation of the Columbia object image library dataset [24]. | Object |
| Caltech101-7 | 1474 | 7 | GABOR(48), WM(40), CENT(254), HOG(1984), GIST(512), LBP(928) | This dataset contains 1474 images belonging to seven classes, which are faces, motorbikes, dollar bill, Garfield, stop sign, and windsor chair [4]. | Object |
| Caltech101-20 | 2386 | 20 | GABOR(48), WM(40), CENT(254), HOG(1984), GIST(512), LBP(928) | This is the frequently used subsets of Caltech101 consisting of 20 categories of images built for object recognition tasks [5]. | Object |
| Caltech101-all | 9144 | 102 | GABOR(48), WM(40), CENT(254), HOG(1984), GIST(512), LBP(928) | The Caltech101 dataset contains images from 101 object categories and background (e.g., “helicopter”, “elephant” and “chair” etc.) and a background category that contains the images not from the 101 object categories [8]. | Object |
| CiteSeer | 3312 | 6 | Content(3703), Cites(4732) | The archive contains 3312 documents over the 6 labels (Agents,IR,DB,AI,HCI,ML) [13]. | Text |
| Cora | 2708 | 7 | Content(1433), Inbound(2708), Outbound(2708), Cites(2708) | The archive contains 2708 documents over the 7 labels (Neural_Networks, Rule_Learning, Reinforcement_Learning, Probabilistic_Methods, Theory, Genetic_Algorithms,Case_Based) [13]. | Text |
| Handwritten | 2000 | 10 | FOU(76), FAC(216), KAR(64), PIX(240), ZER(47), MOR(6) | This dataset consists of features of handwritten numerals ('0'--'9') extracted from a collection of Dutch utility maps [7]. | Image |
| MNIST-10k | 10000 | 10 | ISO(30), LDA(9), NPE(30) | A freely available and well-known handwritten database for image recognition consisting of four categories from digit 0 to digit 9, and each category has 1,000 samples evenly [1]. | Handwritten |
| MNIST-4 | 4000 | 4 | ISO(30), LDA(9), NPE(30) | MNIST-4 is a subset of MNIST-10k consisting of four categories from digit 0 to digit 3 [5]. | Handwritten |
| Movies | 617 | 17 | Keywords(1878), Actors(1398) | It is a movie corpus extracted from IMDb. [22] | Text |
| MSRC-v5 | 210 | 7 | CM(24), HOG(576), GIST(512), LBP(256), CENT(254) | A subset of the Microsoft Research in Cambridge dataset [21]. | Image |
| NUS-WIDE-OBJ | 30000 | 31 | CH(65), CM(226), CORR(145), EDH(74), WT(129) | It is a dataset for object recognition which consists of 30000 images in 31 classes[9]. | Image |
| NUS-WIDE | 2400 | 12 | CH(64), CORR(144), EDH(75), WT(128), CM55(225) | 12 categories of animal images selected from the NUS-WIDE-OBJ dataset, and the first 200 images are selected for each category [3]. | Object |
| ORL | 400 | 40 | View1(4096), View2(3304), View3(6750) | Face dataset contains 400 images of 40 distinct subjects. For each category, images were taken at different times, lights, facial expressions (open / closed eyes, smiling or not) and facial details (with glasses / without glasses) [23] . | Face |
| OutdoorScene | 2688 | 8 | GIST(512), HOG(432), LBP(256), Gabor(48) | The dataset has 2688 outdoor scene images consisting of 8 groups [13]. | Object |
| ProteinFold | 694 | 27 | View1(27), View2(27), View3(27), View4(27), View5(27), View6(27), View7(27), View8(27), View9(27), View10(27), View11(27), View12(27) | Multiply kernel learning dataset on protein fold prediction [19]. | Protein |
| Prokaryotic | 551 | 4 | gene-repert(393), proteome-comp(3), Text(438) | It contains prokaryotic species described with heterogeneous multi-view data including textual data and different genomic representations.[20] | Genome |
| Reuters-1200 | 1200 | 6 | English(2000), French(2000), German(2000), Spanish(2000), Italian(2000) | It contains 6 samples of 1200 documents over 6 labels, and desribed by 5 views of 2000 words each [15]. | Text |
| Reuters-1500 | 1500 | 6 | English(21531), French(24893), German(34279), Spanish(15506), Italian(11547) | The dataset is collection of 1500 documents which are expressed in five different languages (Italian, Spanish, French, German and English) and the corresponding translations [14]. | Text |
| Reuters | 18758 | 6 | English(21531), French(24893), German (34279), Spanish (15506), Italian(11547) | This dataset consists of documents that are written in five different languages and their translations[9]. | Text |
| UCI | 2000 | 10 | Intensity(240), FOU(76), MOR(6) | Obtained from the UCI machine learning repository, this dataset consists of 2000 handwritten digit images belonging to 10 digits with each digit containing 200 samples [4]. | Handwritten |
| WebKB | 1051 | 2 | Anchor(1,840), Content(3,000) | The dataset for web pages collected from computer science department web sites at four universities: Cornell University, University of Washington,University of Wisconsin, and University of Texas [6]. | Text |
| Wikipedia | 2866 | 10 | Word(128), SIFT(10) | Wikipedia dataset is the most widely-used dataset for cross-media retrieval. It is based on Wikipedia’s "featured articles", a continually updated article collection [11]. | Text |
| Wikipedia-test | 693 | 10 | Word(128), SIFT(10) | The test set of the Wikipedia collection [10]. | Text |
| Yale | 165 | 15 | Intensity(4096), LBP(3304), Gabor(6750) | This dataset consists of 165 gray-scale face images belonging to 15 subjects with each subject containing 11 images [4]. | Face |
- FOU: Fourier coefficients of the character shapes
- FAC: Profile correlations
- PIX: Pixel averages in 2 × 3 windows
- ZER: Zernike moment
- MOR: Morphological features.
- Gabor: Gabor feature
- WM: Wavelet moments
- CENT: CENTRIST feature [17]
- HOG: Histogram of oriented gradients feature
- GIST: GIST feature [18]
- LBP: Local binary patterns feature
- CH: Color Histogram
- TH: Texture Histogram
- CM: Color moments
- CS: Color similiarity
- CORR: Color correlation
- EDH: Edge distribution
- WT: Wavelet texture
- SD: Shape descriptor
- FSM: Fine scale margin
- SIFT: Scale Invariant Feature Transform
[1] L. Deng, “The MNIST Database of Handwritten Digit Images for Machine Learning Research [Best of the Web],” IEEE Signal Processing Magazine, vol. 29, no. 6, pp. 141–142, Jan. 2012, doi: 10.1109/MSP.2012.2211477.
[2] J. Winn and N. Jojic, “LOCUS: learning object classes with unsupervised segmentation,” in Tenth IEEE International Conference on Computer Vision (ICCV’05) Volume 1, Oct. 2005, pp. 756-763 Vol. 1. doi: 10.1109/ICCV.2005.148.
[3] B. Wang, Y. Xiao, Z. Li, X. Wang, X. Chen, and D. Fang, “Robust Self-Weighted Multi-View Projection Clustering,” in Proc. AAAI Conf. Artif. Intell., Apr. 2020, pp. 6110–6117. doi: 10.1609/aaai.v34i04.6075.
[4] M.-S. Chen, C.-D. Wang, and J.-H. Lai, “Low-rank Tensor Based Proximity Learning for Multi-view Clustering,” IEEE Transactions on Knowledge and Data Engineering, pp. 1–1, Jan. 2022, doi: 10.1109/TKDE.2022.3151861.
[5] X. Li, H. Zhang, R. Wang, and F. Nie, “Multiview Clustering: A Scalable and Parameter-Free Bipartite Graph Fusion Method,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 44, no. 1, pp. 330–344, Jan. 2022, doi: 10.1109/TPAMI.2020.3011148.
[6] Q. Qiang, B. Zhang, F. Wang, and F. Nie, “Fast Multi-view Discrete Clustering with Anchor Graphs,” in Proc. AAAI Conf. Artif. Intell., May 2021, pp. 9360–9367. doi: 10.1609/aaai.v35i11.17128.
[7] F. Nie, L. Tian, and X. Li, “Multiview clustering via adaptively weighted procrustes,” in Proc. ACM Int. Conf. Knowl. Discov. Data Min., 2018, pp. 2022–2030. doi: 10.1145/3219819.3220049.
[8] Z. Zhang, L. Liu, F. Shen, H. T. Shen, and L. Shao, “Binary Multi-View Clustering,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 41, no. 7, pp. 1774–1782, Jul. 2019, doi: 10.1109/TPAMI.2018.2847335.
[9] Y. Li, F. Nie, H. Huang, and J. Huang, “Large-scale multi-view spectral clustering via bipartite graph,” in Proc. AAAI Conf. Artif. Intell., 2015. doi: 10.1609/aaai.v29i1.9598.
[10] S. Shi, F. Nie, R. Wang, and X. Li, “Fast Multi-View Clustering via Prototype Graph,” IEEE Transactions on Knowledge and Data Engineering, vol. 35, no. 1, pp. 443–455, Jan. 2023, doi: 10.1109/TKDE.2021.3078728.
[11] N. Rasiwasia et al., “A new approach to cross-modal multimedia retrieval,” in Proc. ACM Int. Conf. Multimedia, 2010, pp. 251–260. doi: 10.1145/1873951.1873987.
[12] C. H. Lampert, H. Nickisch, and S. Harmeling, “Learning to detect unseen object classes by between-class attribute transfer,” in Proc. IEEE Conf. Comput. Vis. Pattern Recognit., Jun. 2009, pp. 951–958. doi: 10.1109/CVPR.2009.5206594.
[13] S.-G. Fang, D. Huang, X.-S. Cai, C.-D. Wang, C. He, and Y. Tang, “Efficient Multi-view Clustering via Unified and Discrete Bipartite Graph Learning,” IEEE Transactions on Neural Networks and Learning Systems, pp. 1–12, Apr. 2023, doi: 10.1109/TNNLS.2023.3261460.
[14] H. Wei, L. Chen, C. L. P. Chen, J. Duan, R. Han, and L. Guo, “Fuzzy clustering for multiview data by combining latent information,” Applied Soft Computing, vol. 126, p. 109140, Sep. 2022, doi: 10.1016/j.asoc.2022.109140.
[15] S. Huang, Z. Kang, and Z. Xu, “Auto-weighted multi-view clustering via deep matrix decomposition,” Pattern Recognition, vol. 97, p. 107015, Jan. 2020, doi: 10.1016/j.patcog.2019.107015.
[16] J. Yang and C.-T. Lin, “Multi-View Adjacency-Constrained Hierarchical Clustering,” IEEE Transactions on Emerging Topics in Computational Intelligence, pp. 1–13, 2022, doi: 10.1109/TETCI.2022.3221491.
[17] J. Wu and J. M. Rehg, “Centrist: A visual descriptor for scene categorization,” IEEE transactions on pattern analysis and machine intelligence, vol. 33, no. 8, pp. 1489–1501, 2010.
[18] A. Oliva and A. Torralba, “Modeling the Shape of the Scene: A Holistic Representation of the Spatial Envelope,” International Journal of Computer Vision, vol. 42, no. 3, pp. 145–175, May 2001, doi: 10.1023/A:1011139631724.
[19] X. Liu et al., “Efficient and Effective Regularized Incomplete Multi-View Clustering,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 43, no. 8, pp. 2634–2646, Aug. 2021, doi: 10.1109/TPAMI.2020.2974828.
[20] X. Niu, C. Zhang, X. Zhao, L. Hu, and J. Zhang, “A multi-view ensemble clustering approach using joint affinity matrix,” Expert Systems with Applications, vol. 216, p. 119484, Apr. 2023, doi: 10.1016/j.eswa.2022.119484.
[21] W. Xia, Q. Gao, Q. Wang, X. Gao, C. Ding, and D. Tao, “Tensorized Bipartite Graph Learning for Multi-View Clustering,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 45, no. 4, pp. 5187–5202, Apr. 2023, doi: 10.1109/TPAMI.2022.3187976.
[22] X. Xie and Y. Xiong, “Generalized multi-view learning based on generalized eigenvalues proximal support vector machines,” Expert Systems with Applications, vol. 194, p. 116491, May 2022, doi: 10.1016/j.eswa.2021.116491.
[23] S. Luo, C. Zhang, W. Zhang, and X. Cao, “Consistent and Specific Multi-View Subspace Clustering,” in Proc. AAAI Conf. Artif. Intell., Apr. 2018. doi: 10.1609/aaai.v32i1.11617.
[24] J. Wu, Z. Lin, and H. Zha, “Essential Tensor Learning for Multi-View Spectral Clustering,” IEEE Transactions on Image Processing, vol. 28, no. 12, pp. 5910–5922, Feb. 2019, doi: 10.1109/TIP.2019.2916740.
[25] G.-Y. Zhang, D. Huang, and C.-D. Wang, “Facilitated low-rank multi-view subspace clustering,” Knowledge-Based Systems, vol. 260, p. 110141, Jan. 2023, doi: 10.1016/j.knosys.2022.110141.