Word embedding based on low-rank doubly stochastic matrix decomposition

 |  Login

Show simple item record

dc.contributor Aalto-yliopisto fi
dc.contributor Aalto University en
dc.contributor.author Sedov, Denis
dc.contributor.author Yang, Zhirong
dc.date.accessioned 2019-01-14T09:23:51Z
dc.date.available 2019-01-14T09:23:51Z
dc.date.issued 2018
dc.identifier.citation Sedov , D & Yang , Z 2018 , Word embedding based on low-rank doubly stochastic matrix decomposition . in Neural Information Processing : 25th International Conference, ICONIP 2018 Siem Reap, Cambodia, December 13–16, 2018 Proceedings, Part III . vol. 3 , Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) , vol. 11303 LNCS , SPRINGER , pp. 90-100 , International Conference on Neural Information Processing , Siem Reap , Cambodia , 13/12/2018 . en
dc.identifier.isbn 978-3-030-04181-6
dc.identifier.isbn 978-3-030-04182-3
dc.identifier.issn 0302-9743
dc.identifier.issn 1611-3349
dc.identifier.other PURE UUID: b778e605-84ee-4a02-9e7d-2c3cb56a196c
dc.identifier.other PURE ITEMURL: https://research.aalto.fi/en/publications/word-embedding-based-on-lowrank-doubly-stochastic-matrix-decomposition(b778e605-84ee-4a02-9e7d-2c3cb56a196c).html
dc.identifier.other PURE LINK: https://www.springerprofessional.de/word-embedding-based-on-low-rank-doubly-stochastic-matrix-decomp/16310380
dc.identifier.uri https://aaltodoc.aalto.fi/handle/123456789/36010
dc.description.abstract Word embedding, which encodes words into vectors, is an important starting point in natural language processing and commonly used in many text-based machine learning tasks. However, in most current word embedding approaches, the similarity in embedding space is not optimized in the learning. In this paper we propose a novel neighbor embedding method which directly learns an embedding simplex where the similarities between the mapped words are optimal in terms of minimal discrepancy to the input neighborhoods. Our method is built upon two-step random walks between words via topics and thus able to better reveal the topics among the words. Experiment results indicate that our method, compared with another existing word embedding approach, is more favorable for various queries. en
dc.format.extent 10
dc.format.extent 90-100
dc.language.iso en en
dc.publisher Springer
dc.relation.ispartof International Conference on Neural Information Processing en
dc.relation.ispartofseries Neural Information Processing en
dc.relation.ispartofseries Volume 3 en
dc.relation.ispartofseries Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) en
dc.relation.ispartofseries Volume 11303 LNCS en
dc.rights embargoedAccess en
dc.subject.other 113 Computer and information sciences en
dc.title Word embedding based on low-rank doubly stochastic matrix decomposition en
dc.type A4 Artikkeli konferenssijulkaisussa fi
dc.description.version Peer reviewed en
dc.contributor.department Professorship Kaski S.
dc.contributor.department Probabilistic Machine Learning
dc.contributor.department Department of Computer Science en
dc.subject.keyword Nonnegative matrix factorization
dc.subject.keyword Word embedding
dc.subject.keyword Cluster analysis
dc.subject.keyword Doubly stochastic
dc.subject.keyword 113 Computer and information sciences
dc.identifier.urn URN:NBN:fi:aalto-201901141193
dc.date.embargo info:eu-repo/date/embargoEnd/2019-12-01


Files in this item

Files Size Format View

There are no files associated with this item.

This item appears in the following Collection(s)

Show simple item record

Search archive


Advanced Search

article-iconSubmit a publication

Browse

My Account