AutoKGE: Towards Automated Knowledge Graph Embedding

PhD Thesis Proposal Defence


Title: "AutoKGE: Towards Automated Knowledge Graph Embedding"

by

Mr. Yongqi ZHANG


Abstract:

Knowledge graph (KG) embedding aims to encode the entities and relations 
in KG into low dimensional vector space while preserving the inherent 
structure of KG. To learn better embeddings, three aspects of KG area, 
i.e., negative sampling, semantic information and structural information, 
are extensively studied in the literature. However, since different KGs 
have complex and distinct patterns, a single model is usually hard to 
adapt well to different KGs. Recently, automated machine learning (AutoML) 
has exhibited its power in many machine learning tasks. Inspired by the 
recent success of AutoML in both academia and industry, we propose AutoKGE 
in this thesis to address the three KG problems in different ways. AutoKGE 
is not only new to the literature, but also opens up new directions in 
analyzing and designing the KG embedding models. In detail, we propose 
AutoNS to automatically keep track of the dynamic distribution of negative 
triplets. The proposed negative sampling algorithm is simple but very 
efficient in sampling high-quality negative triplets. To capture the 
different semantic information, we propose AutoSF to automatically design 
SFs for distinct KGs regarding the relation patterns. Novel, better and 
KG-dependent scoring functions are designed through our algorithm. To 
explore the structure information, we propose S2E to distill structural 
information and combine it with semantic information based on the 
relational path. As a neural architecture search (NAS) problem, the 
searched models adaptively combine the structural and semantic information 
in various KG tasks. Extensive experiments demonstrate the effectiveness 
of the searched models and efficiency of each search algorithms.


Date:			Monday, 9 December 2019

Time:                  	2:00pm - 4:00pm

Venue:                  Room 2132B
                         (lift 19)

Committee Members:	Prof. Lei Chen (Supervisor)
 			Prof. Dik-Lun Lee (Chairperson)
 			Dr. Qifeng Chen
 			Prof. Nevin Zhang


**** ALL are Welcome ****