Ensemble Solutions for Link-Prediction in Knowledge Graphs

Slides:



Advertisements
Similar presentations
1 Understanding User Roles Understanding User Roles ( in ontologizing the Ontolog body of knowledge) Lisa Dawn Colvin April 20, 2006.
Advertisements

Autonomic Scaling of Cloud Computing Resources
Date: 2014/05/06 Author: Michael Schuhmacher, Simon Paolo Ponzetto Source: WSDM’14 Advisor: Jia-ling Koh Speaker: Chen-Yu Huang Knowledge-based Graph Document.
Probabilistic Latent-Factor Database Models Denis Krompaß 1, Xueyan Jiang 1,Maximilian Nickel 2 and Volker Tresp 1,3 1 Department of Computer Science.
COLLABORATIVE FILTERING Mustafa Cavdar Neslihan Bulut.
Non-Negative Tensor Factorization with RESCAL Denis Krompaß 1, Maximilian Nickel 1, Xueyan Jiang 1 and Volker Tresp 1,2 1 Department of Computer Science.
Copyright 2005 Northrop Grumman Corporation 0 Critical Success Factors for system-of-system architecture / engineering 25 October 2006 Neil Siegel Sector.
MCS 2005 Round Table In the context of MCS, what do you believe to be true, even if you cannot yet prove it?
Information Agents for Autonomous Acquisition of Sensor Network Data A. Rogers and N. R. Jennings University of Southampton, UK M. A. Osborne and S. J.
Heterogeneous Consensus Learning via Decision Propagation and Negotiation Jing Gao† Wei Fan‡ Yizhou Sun†Jiawei Han† †University of Illinois at Urbana-Champaign.
Semi-Supervised Clustering Jieping Ye Department of Computer Science and Engineering Arizona State University
Ranking by Odds Ratio A Probability Model Approach let be a Boolean random variable: document d is relevant to query q otherwise Consider document d as.
Large-Scale Factorization of Type- Constrained Multi-Relational Data Denis Krompaß 1, Maximilian Nickel 2 and Volker Tresp 1,3 1 Department of Computer.
Representation learning for Knowledge Bases LivesIn BornIn LocateIn Friendship Nationality Nicole Kidman PerformIn Nationality Sydney Hugh Jackman Australia.
CS157A Spring 05 Data Mining Professor Sin-Min Lee.
1 Collaborative Filtering: Latent Variable Model LIU Tengfei Computer Science and Engineering Department April 13, 2011.
Predicting Missing Provenance Using Semantic Associations in Reservoir Engineering Jing Zhao University of Southern California Sep 19 th,
OMAP: An Implemented Framework for Automatically Aligning OWL Ontologies SWAP, December, 2005 Raphaël Troncy, Umberto Straccia ISTI-CNR
Data Mining Techniques
Cao et al. ICML 2010 Presented by Danushka Bollegala.
Machine Learning1 Machine Learning: Summary Greg Grudic CSCI-4830.
Integrating Neural Network and Genetic Algorithm to Solve Function Approximation Combined with Optimization Problem Term presentation for CSC7333 Machine.
Distributed Computing Rik Sarkar. Distributed Computing Old style: Use a computer for computation.
Iowa State University Department of Computer Science Center for Computational Intelligence, Learning, and Discovery Harris Lin, Neeraj Koul, and Vasant.
Towards an ecosystem of data and ontologies Mathieu d’Aquin and Enrico Motta Knowledge Media Institute The Open University.
Biswanath Panda, Mirek Riedewald, Daniel Fink ICDE Conference 2010 The Model-Summary Problem and a Solution for Trees 1.
Machine Learning.
LATENT SEMANTIC INDEXING Hande Zırtıloğlu Levent Altunyurt.
Keyword Query Routing.
CS157B Fall 04 Introduction to Data Mining Chapter 22.3 Professor Lee Yu, Jianji (Joseph)
Natural Language Questions for the Web of Data Mohamed Yahya 1, Klaus Berberich 1, Shady Elbassuoni 2 Maya Ramanath 3, Volker Tresp 4, Gerhard Weikum 1.
Querying Factorized Probabilistic Triple Databases Denis Krompaß 1, Maximilian Nickel 2 and Volker Tresp 1,3 1 Department of Computer Science. Ludwig Maximilian.
Natural Language Questions for the Web of Data 1 Mohamed Yahya, Klaus Berberich, Gerhard Weikum Max Planck Institute for Informatics, Germany 2 Shady Elbassuoni.
Data Mining BY JEMINI ISLAM. Data Mining Outline: What is data mining? Why use data mining? How does data mining work The process of data mining Tools.
Natural Language Questions for the Web of Data Mohamed Yahya, Klaus Berberich, Gerhard Weikum Max Planck Institute for Informatics, Germany Shady Elbassuoni.
Iana Atanassova Research: – Information retrieval in scientific publications exploiting semantic annotations and linguistic knowledge bases – Ranking algorithms.
Instructor: Pedro Domingos
Copyright © 2001, SAS Institute Inc. All rights reserved. Data Mining Methods: Applications, Problems and Opportunities in the Public Sector John Stultz,
Using Bayesian Networks to Predict Plankton Production from Satellite Data By: Rob Curtis, Richard Fenn, Damon Oberholster Supervisors: Anet Potgieter,
Higher Computing Science 2016 Prelim Revision. Topics to revise Computational Constructs parameter passing (value and reference, formal and actual) sub-programs/routines,
Advanced Gene Selection Algorithms Designed for Microarray Datasets Limitation of current feature selection methods: –Ignores gene/gene interaction: single.
Enhanced hypertext categorization using hyperlinks Soumen Chakrabarti (IBM Almaden) Byron Dom (IBM Almaden) Piotr Indyk (Stanford)
Instance Discovery and Schema Matching With Applications to Biological Deep Web Data Integration Tantan Liu, Fan Wang, Gagan Agrawal {liut, wangfa,
1 Intelligent Information System Lab., Department of Computer and Information Science, Korea University Semantic Social Network Analysis Kyunglag Kwon.
A Review of Relational Machine Learning for Knowledge Graphs CVML Reading Group Xiao Lin.
Multi-Modal Bayesian Embeddings for Learning Social Knowledge Graphs Zhilin Yang 12, Jie Tang 1, William W. Cohen 2 1 Tsinghua University 2 Carnegie Mellon.
Ke (Kevin) Wu1,2, Philip Watters1, Malik Magdon-Ismail1
Brief Intro to Machine Learning CS539
Ensembling Diverse Approaches to Question Answering
Semi-Supervised Clustering
Instructor: Pedro Domingos
CSEP 546 Data Mining Machine Learning
Machine Learning Ali Ghodsi Department of Statistics
Associative Query Answering via Query Feature Similarity
به نام خدا Big Data and a New Look at Communication Networks Babak Khalaj Sharif University of Technology Department of Electrical Engineering.
CSEP 546 Data Mining Machine Learning
CIKM Competition 2014 Second Place Solution
Variational Knowledge Graph Reasoning
CIKM Competition 2014 Second Place Solution
Semantic Network & Knowledge Graph
Knowledge Base Completion
Property consolidation for entity browsing
Chap. 7 Regularization for Deep Learning (7.8~7.12 )
Information Networks: State of the Art
Enriching Taxonomies With Functional Domain Knowledge
Predicting Loan Defaults
Dan Roth Department of Computer Science
Bug Localization with Combination of Deep Learning and Information Retrieval A. N. Lam et al. International Conference on Program Comprehension 2017.
Embedding based entity summarization
Heterogeneous Graph Convolutional Network
Presentation transcript:

Ensemble Solutions for Link-Prediction in Knowledge Graphs Denis Krompaß1,2 and Volker Tresp1,2 1 Department of Computer Science. Ludwig Maximilian University, 2 Corporate Technology, Siemens AG 12.09.2015

Outline Knowledge Graphs, what are they and what are they good for? Representation Learning in Knowledge Graphs State of the Art Latent Variable Models Integrating Prior Knowledge about Relation-Types Analyzing the Complementary “Potential” of State of the Art Representation Learning Algorithms

Knowledge Graphs Stores facts about the world as relations between entities. Entities are no longer just strings but real world objects with attributes, taxonomic information and relations to other objects. (AlbertEinstein, bornIn, Ulm) Providing a machine with semantic information: Search engines Information retrieval Word-sense disambiguation … Prominent Examples: Google Knowledge Graph IBM Watson

Learning in Knowledge Graphs Latent Variable Model Knowledge Graph Triples bornIn Similarities 0.2 0.9 1.2 0.3 0.7 0.9 Albert Einstein -0.4 -0.3 -0.8 -0.5 -0.4 1.3 -1.1 0.1 1.3 1.3 bornIn 0.3 2.1 0.7 0.3 0.3 0.1 0.6 -0.9 0.2 0.1 ULM -0.1 1.7 -0.2 -0.1 Latent representations (or embeddings) for Entities and Relation-Types that disentangle complex relationships observed in the data (semantics). Link-Prediction Link-based Clustering Disambiguation

State of the Art Latent Variable Models RESCAL Third-Order Tensor Factorization Methods Least-Squares Cost Function TransE Distance-based Method Ranking Cost Function Google Knowledge Vault Multi-way Neural Network (mwNN) Logistic Cost Function Problem: Large Knowledge Graphs Contain Millions of Entities and thousands of Relation-Types Low dimensional representations have to be learned Try to find ways to increase prediction-quality under this constraint

Prior Knowledge about Relation-Types Type-Constraints (From the Schema) Local closed-world assumption (From the Data) Domain and Range Constraints for Relation-Types Integration in model training RESCAL TransE Google KVault Neural Network With low-dimensional embeddings Latent Variable Model Prediction of new triples +40% (YAGO2*) +77% (Freebase*) Denis Krompaß, Stephan Baier and Volker Tresp. Type-Constrained Representation Learning in Knowledge Graphs. 14th International Semantic Web Conference (ISWC), 2015 *Results on large samples from these knowledge graphs Link-Prediction Improvement +54% (DBpedia-Music*)

Complementary Prediction? State of the art models differ in many aspects Diverse predictors Analysis to which degree the models are complementary Combine through arithmetic mean Use Plat scaling for mapping the different outputs to probabilities 70% Training Set 10% Validation Set Hyperparameter Tuning + Plat Scaling 20 % Test Set

Results Ensemble has always much better link-prediction quality

Results Ensemble has always much better link-prediction quality Best complement is between TransE and mwNN

Results Ensemble has always much better link-prediction quality Best complement is between TransE and mwNN RESCAL provides only complementary predictions in case of the Freebase dataset

Results Ensemble has always much better link-prediction quality Best complement is between TransE and mwNN RESCAL provides only complementary predictions in case of the Freebase dataset For the local closed-world assumption, very similar observations could be made

Summary Models are complementary to each other This applies especially when low dimensional embeddings are used (d=10) Ensemble with d=10 comparable to best single predictor with d=100 Up to more than 10% improvement on top of the improvements achieved when Type-Constraints or the Local closed-world assumption are exploited

Questions ? http://www.dbs.ifi.lmu.de/~krompass/ Denis.Krompass@siemens.com