{"type":"ScholarlyArticle","title":"Quaternion Ordinal Embedding","authors":[{"name":"Wenzheng Hou","slug":"wenzheng-hou","given":"Wenzheng","family":"Hou","affiliation":"Key Laboratory of Intelligent Information Processing , Institute of Computing Technology , CAS; School of Computer Science and Technology , University of Chinese Academy of Sciences"},{"name":"Qianqian Xu","slug":"qianqian-xu","given":"Qianqian","family":"Xu","affiliation":"Key Laboratory of Intelligent Information Processing , Institute of Computing Technology , CAS"},{"name":"Ke Ma","slug":"ke-ma","given":"Ke","family":"Ma","affiliation":"School of Computer Science and Technology , University of Chinese Academy of Sciences"},{"name":"Qianxiu Hao","slug":"qianxiu-hao","given":"Qianxiu","family":"Hao","affiliation":"Key Laboratory of Intelligent Information Processing , Institute of Computing Technology , CAS; School of Computer Science and Technology , University of Chinese Academy of Sciences"},{"name":"Qingming Huang","slug":"qingming-huang","given":"Qingming","family":"Huang","affiliation":"Key Laboratory of Intelligent Information Processing , Institute of Computing Technology , CAS; School of Computer Science and Technology , University of Chinese Academy of Sciences; Key Laboratory of Big Data Mining and Knowledge Management , Chinese Academy of Sciences; Artificial Intelligence Research Center , Peng Cheng Laboratory"}],"abstract":"Ordinal embedding (OE) aims to project objects into a low-dimensional space while preserving their ordinal constraints as well as possible. Generally speaking, a reasonable OE algorithm should simultaneously capture  a) semantic meaning and b) the ordinal relationship of the objects. However, most of the existing methods merely focus on b). To address this issue, our goal in this paper is to seek a generic OE method to embrace the two features simultaneously. We argue that different dimensions of vector-based embedding are naturally entangled with each other. To realize a), we expect to decompose the D dimensional embedding space into D different semantic subspaces, where each subspace is associated with a  matrix representation. Unfortunately, introducing a matrix-based representation requires far more complex parametric space than its vector-based counterparts. Thanks to the algebraic property of quaternions, we are able to find a more efficient way to represent a matrix with quaternions. For b), inspired by the classic chordal Grassmannian distance, a new distance function is defined to measure the distance between different quaternions/matrices, on top of which we construct a generic OE loss function. Experimental results for different tasks on both simulated and real-world datasets verify the effectiveness of our proposed method.","doi":"10.24963/ijcai.2022/287","url":"https://ijcai.org/proceedings/2022/287","pdf":"https://ijcai.org/proceedings/2022/0287.pdf","markdown":"https://ijcai.org/proceedings/2022/287.md","bibtex":"https://ijcai.org/proceedings/2022/bibtex/287","volume":{"slug":"2022","title":"Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence","short_title":"IJCAI 2022","url":"https://ijcai.org/proceedings/2022/","year":2022,"location":"Vienna","dates":"23-29 July 2022","editors":"Luc De Raedt","isbn":"978-1-956792-00-3","publisher":"International Joint Conferences on Artificial Intelligence Organization","doi_prefix":"10.24963/ijcai.2022"},"section":"Data Mining","track":"Main Track","number":287,"pages":{"first":2066,"last":2072},"keywords":["Data Mining: Applications"],"videos":[{"label":"Video #1","url":"https://slideslive.com/38986486","length_sec":65,"provider":"slideslive"},{"label":"Video #2","url":"https://slideslive.com/38984629","length_sec":259,"provider":"slideslive"}],"references":[{"position":1,"raw":"[Agarwal et al., 2007] Sameer Agarwal, Josh Wills, Lawrence Cayton, Gert Lanckriet, David Kriegman, and Serge Belongie. Generalized non-metric multidimensional scaling. In AIS, pages 11-18, 2007.","title":"Generalized non-metric multidimensional scaling","authors":"Sameer Agarwal, Josh Wills, Lawrence Cayton, Gert Lanckriet, David Kriegman, Serge Belongie","year":"2007","venue":"AIS","doi":null,"url":null,"internal_url":null},{"position":2,"raw":"[Bar-Itzhack, 2000] Itzhack Y Bar-Itzhack. New method for extracting the quaternion from a rotation matrix. JGCD, 23:1085-1087, 2000.","title":"New method for extracting the quaternion from a rotation matrix","authors":"Y Itzhack, Bar-Itzhack","year":"2000","venue":"JGCD","doi":null,"url":null,"internal_url":null},{"position":3,"raw":"[Bengtsson et al., 2007] Ingemar Bengtsson, Wojciech Bruzda, Åsa Ericsson, Jan-Åke Larsson, Wojciech Tadej, and Karol Życzkowski. Mutually unbiased bases and hadamard matrices of order six. J Math Phys, 48:052106, 2007.","title":"Mutually unbiased bases and hadamard matrices of order six","authors":"Ingemar Bengtsson, Wojciech Bruzda, Åsa Ericsson, Jan-Åke Larsson, Wojciech Tadej, Karol Życzkowski","year":"2007","venue":"J Math Phys","doi":null,"url":null,"internal_url":null},{"position":4,"raw":"[Kingma and Ba, 2015] Diederik P Kingma and Jimmy Ba. Adam: A method for stochastic optimization. In ICLR, 2015.","title":"Adam: A method for stochastic optimization","authors":"P Diederik, Jimmy Kingma, Ba","year":"2015","venue":"ICLR","doi":null,"url":null,"internal_url":null},{"position":5,"raw":"[Kruskal, 1964a] Joseph B Kruskal. Multidimensional scal- ing by optimizing goodness of fit to a nonmetric hypothe- sis. Psychometrika, 29:1-27, 1964.","title":"Multidimensional scaling by optimizing goodness of fit to a nonmetric hypothesis","authors":"Joseph B Kruskal, Kruskal","year":"1964","venue":"Psychometrika","doi":null,"url":null,"internal_url":null},{"position":6,"raw":"[Kruskal, 1964b] Joseph B Kruskal. Nonmetric multidi- mensional scaling: a numerical method. Psychometrika, 29:115-129, 1964.","title":"Nonmetric multidimensional scaling: a numerical method","authors":"Joseph B Kruskal, Kruskal","year":"1964","venue":"Psychometrika","doi":null,"url":null,"internal_url":null},{"position":7,"raw":"[Liu et al., 2011] Yiguang Liu, Shuzhi Sam Ge, Chunguang Li, and Zhisheng You. k-ns: A classifier by the distance to the nearest subspace. IEEE Trans, 22:1256-1268, 2011.","title":"k-ns: A classifier by the distance to the nearest subspace","authors":"Liu","year":"2011","venue":"IEEE Trans","doi":null,"url":null,"internal_url":null},{"position":8,"raw":"[Luce, 2012] R Duncan Luce. Individual choice behavior: A theoretical analysis. Courier Corporation, 2012.","title":"Individual choice behavior: A theoretical analysis","authors":"Luce Duncan","year":"2012","venue":null,"doi":null,"url":null,"internal_url":null},{"position":9,"raw":"[Ma et al., 2019] Ke Ma, Qianqian Xu, Zhiyong Yang, and Xiaochun Cao. Less but better: Generalization enhance- ment of ordinal embedding via distributional margin. In AAAI, pages 2978-2985, 2019.","title":"Less but better: Generalization enhancement of ordinal embedding via distributional margin","authors":"Ke Ma, Qianqian Xu, Zhiyong Yang, Xiaochun Cao","year":"2019","venue":"AAAI","doi":null,"url":null,"internal_url":null},{"position":10,"raw":"[Markley, 2008] F Landis Markley. Unit quaternion from ro- tation matrix. JGCD, 31:440-442, 2008.","title":"Unit quaternion from rotation matrix","authors":"Markley Landis","year":"2008","venue":"JGCD","doi":null,"url":null,"internal_url":null},{"position":11,"raw":"[McFee and Lanckriet, 2011] Brian McFee and Gert Lanck- riet. Learning multi-modal similarity. JMLR, 12:491-523, 2011.","title":"Learning multi-modal similarity","authors":"Brian Mcfee, Gert Lanckriet","year":"2011","venue":"JMLR","doi":null,"url":null,"internal_url":null},{"position":12,"raw":"[Moghaddam and Pentland, 1997] Baback Moghaddam and Alex Pentland. Probabilistic visual learning for object rep- resentation. IEEE TPAMI, 19:696-710, 1997.","title":"Probabilistic visual learning for object representation","authors":"Pentland Moghaddam, Baback Moghaddam, Alex Pentland","year":"1997","venue":"IEEE TPAMI","doi":null,"url":null,"internal_url":null},{"position":13,"raw":"[Nishiyama et al., 2005] Masashi Nishiyama, Osamu Yam- aguchi, and Kazuhiro Fukui. Face recognition with the multiple constrained mutual subspace method. In AVBPA, pages 71-80, 2005.","title":"Face recognition with the multiple constrained mutual subspace method","authors":"Masashi Nishiyama, Osamu Yamaguchi, Kazuhiro Fukui","year":"2005","venue":"AVBPA","doi":null,"url":null,"internal_url":null},{"position":14,"raw":"[Pang et al., 2009] Y. Pang, Y. Yuan, and X. Li. Iterative sub- space analysis based on feature line distance. IEEE TIP, 18:903-907, 2009.","title":"Iterative subspace analysis based on feature line distance","authors":"Y Pang, Y Yuan, X Li","year":"2009","venue":"IEEE TIP","doi":null,"url":null,"internal_url":null},{"position":15,"raw":"[Shaw and Jebara, 2009] Blake Shaw and Tony Jebara. Structure preserving embedding. In ICML, pages 937- 944, 2009.","title":"Structure preserving embedding","authors":"Jebara Shaw, Blake Shaw, Tony Jebara","year":"2009","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":16,"raw":"[Shepard, 1962a] Roger N Shepard. The analysis of proxim- ities: multidimensional scaling with an unknown distance function. i. Psychometrika, 27:125-140, 1962.","title":"The analysis of proximities: multidimensional scaling with an unknown distance function","authors":"Roger N Shepard, Shepard","year":"1962","venue":"i. Psychometrika","doi":null,"url":null,"internal_url":null},{"position":17,"raw":"[Shepard, 1962b] Roger N Shepard. The analysis of proxim- ities: Multidimensional scaling with an unknown distance function. ii. Psychometrika, 27:219-246, 1962.","title":"The analysis of proximities: Multidimensional scaling with an unknown distance function","authors":"Roger N Shepard, Shepard","year":"1962","venue":"ii. Psychometrika","doi":null,"url":null,"internal_url":null},{"position":18,"raw":"[Sun and Cheng, 2006] Xichen Sun and Qiansheng Cheng. On subspace distance. In ICIAR, pages 81-89, 2006.","title":"Xichen Sun and Qiansheng Cheng. On subspace distance","authors":"Cheng Sun","year":"2006","venue":"ICIAR","doi":null,"url":null,"internal_url":null},{"position":19,"raw":"[Suzuki et al., 2019] Atsushi Suzuki, Jing Wang, Feng Tian, Atsushi Nitanda, and Kenji Yamanishi. Hyperbolic ordinal embedding. In ACML, pages 1065-1080, 2019.","title":"Hyperbolic ordinal embedding","authors":"Atsushi Suzuki, Jing Wang, Feng Tian, Atsushi Nitanda, Kenji Yamanishi","year":"2019","venue":"ACML","doi":null,"url":null,"internal_url":null},{"position":20,"raw":"[Tamuz et al., 2011] Omer Tamuz, Ce Liu, Serge Belongie, Ohad Shamir, and Adam Tauman Kalai. Adaptively learn- ing the crowd kernel. In ICML, pages 673-680, 2011.","title":"Adaptively learning the crowd kernel","authors":"Omer Tamuz, Ce Liu, Serge Belongie, Ohad Shamir, Adam Tauman, Kalai","year":"2011","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":21,"raw":"[Terada and Luxburg, 2014] Yoshikazu Terada and Ulrike Luxburg. Local ordinal embedding. In ICML, pages 847- 855, 2014.","title":"Local ordinal embedding","authors":"Luxburg Terada, Yoshikazu Terada, Ulrike Luxburg","year":"2014","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":22,"raw":"[Turk and Pentland, 1991] Matthew A Turk and Alex P Pent- land. Face recognition using eigenfaces. In CVPR, pages 586-587, 1991.","title":"Face recognition using eigenfaces","authors":"A Matthew, Alex P Turk, Pentland","year":"1991","venue":"CVPR","doi":null,"url":null,"internal_url":null},{"position":23,"raw":"[Van Der Maaten and Weinberger, 2012] Laurens Van Der Maaten and Kilian Weinberger. Stochastic triplet embedding. In MLSP, pages 1-6, 2012.","title":"Stochastic triplet embedding","authors":"Van Der Maaten, Laurens Weinberger, Der Van, Kilian Maaten, Weinberger","year":"2012","venue":"MLSP","doi":null,"url":null,"internal_url":null},{"position":24,"raw":"[Wang et al., 2008] Ruiping Wang, Shiguang Shan, Xilin Chen, and Wen Gao. Manifold-manifold distance with application to face recognition based. CVPR, pages 1-8, 2008.","title":"Manifold-manifold distance with application to face recognition based","authors":"Wang","year":"2008","venue":"CVPR","doi":null,"url":null,"internal_url":null},{"position":25,"raw":"[Wang et al., 2016] Jun Wang, Zhaohong Deng, Kup-Sze Choi, Yizhang Jiang, Xiaoqing Luo, Fu-Lai Chung, and Shitong Wang. Distance metric learning for soft subspace clustering in composite kernel space. PR, 52:113-134, 2016.","title":"Distance metric learning for soft subspace clustering in composite kernel space","authors":"Wang","year":"2016","venue":"PR","doi":null,"url":null,"internal_url":null},{"position":26,"raw":"[Wolf and Shashua, 2004] Lior Wolf and Amnon Shashua. Learning over sets using kernel principal angles. JMLR, 4:913-931, 2004.","title":"Learning over sets using kernel principal angles","authors":"Lior Wolf, Amnon Shashua","year":"2004","venue":"JMLR","doi":null,"url":null,"internal_url":null},{"position":27,"raw":"[Zhang et al., 2019] Shuai Zhang, Yi Tay, Lina Yao, and Qi Liu. Quaternion knowledge graph embeddings. In NIPS, pages 2731-2741, 2019.","title":"Quaternion knowledge graph embeddings","authors":"Shuai Zhang, Yi Tay, Lina Yao, Qi Liu","year":"2019","venue":"NIPS","doi":null,"url":null,"internal_url":null},{"position":28,"raw":"[Zuccon et al., 2009] Guido Zuccon, Leif A Azzopardi, and CJ Van Rijsbergen. Semantic spaces: Measuring the dis- tance between different subspaces. In ISQI, pages 225- 236, 2009.","title":"Semantic spaces: Measuring the distance between different subspaces","authors":"Guido Zuccon, Leif A Azzopardi, Van Rijsbergen","year":"2009","venue":"ISQI","doi":null,"url":null,"internal_url":null}],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}