Show simple item record

dc.contributor.authorChen, Hongxu
dc.contributor.authorYin, Hongzhi
dc.contributor.authorWang, Weiqing
dc.contributor.authorWang, Hao
dc.contributor.authorQuoc, Viet Hung Nguyen
dc.contributor.authorLi, Xue
dc.contributor.editorRómer Rosales, Jiliang Tang
dc.date.accessioned2019-05-29T12:44:10Z
dc.date.available2019-05-29T12:44:10Z
dc.date.issued2018
dc.identifier.isbn9781450355520
dc.identifier.doi10.1145/3219819.3219986
dc.identifier.urihttp://hdl.handle.net/10072/379947
dc.description.abstractHeterogenous information network embedding aims to embed heterogenous information networks (HINs) into low dimensional spaces, in which each vertex is represented as a low-dimensional vector, and both global and local network structures in the original space are preserved. However, most of existing heterogenous information network embedding models adopt the dot product to measure the proximity in the low dimensional space, and thus they can only preserve the first-order proximity and are insufficient to capture the global structure. Compared with homogenous information networks, there are multiple types of links (i.e., multiple relations) in HINs, and the link distribution w.r.t relations is highly skewed. To address the above challenging issues, we propose a novel heterogenous information network embedding model PME based on the metric learning to capture both first-order and second-order proximities in a unified way. To alleviate the potential geometrical inflexibility of existing metric learning approaches, we propose to build object and relation embeddings in separate object space and relation spaces rather than in a common space. Afterwards, we learn embeddings by firstly projecting vertices from object space to corresponding relation space and then calculate the proximity between projected vertices. To overcome the heavy skewness of the link distribution w.r.t relations and avoid "over-sampling'' or "under-sampling'' for each relation, we propose a novel loss-aware adaptive sampling approach for the model optimization. Extensive experiments have been conducted on a large-scale HIN dataset, and the experimental results show superiority of our proposed PME model in terms of prediction accuracy and scalability.
dc.description.peerreviewedYes
dc.languageEnglish
dc.publisherAssociation for Computing Machinery (ACM)
dc.publisher.placeUnited States
dc.relation.ispartofchapter42292
dc.relation.ispartofconferencename24th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD)
dc.relation.ispartofconferencetitleKDD'18: PROCEEDINGS OF THE 24TH ACM SIGKDD INTERNATIONAL CONFERENCE ON KNOWLEDGE DISCOVERY & DATA MINING
dc.relation.ispartofdatefrom2018-08-19
dc.relation.ispartofdateto2018-08-23
dc.relation.ispartoflocationLondon, ENGLAND
dc.relation.ispartofpagefrom1177
dc.relation.ispartofpageto1186
dc.subject.fieldofresearchDatabase systems
dc.subject.fieldofresearchcode460505
dc.titlePME: Projected Metric Embedding on Heterogeneous Networks for Link Prediction
dc.typeConference output
dc.type.descriptionE1 - Conferences
dc.type.codeE - Conference Publications
gro.hasfulltextNo Full Text
gro.griffith.authorNguyen, Henry


Files in this item

FilesSizeFormatView

There are no files associated with this item.

This item appears in the following Collection(s)

  • Conference outputs
    Contains papers delivered by Griffith authors at national and international conferences.

Show simple item record