![Neo4j, Inc. All rights reserved 2021
Not that long ago…. Deepmind stated...
“We argue that combinatorial
generalisation must be a top
priority for AI to achieve
human-like abilities, and that
structured representations [i.e.
Graphs] and computations are
key to realizing this objective”](https://image.slidesharecdn.com/awssummitstockholmmay-22final-220519143058-b2aba445/85/ntroducing-to-the-Power-of-Graph-Technology-6-320.jpg)
Uploaded byNeo4j
ntroducing to the Power of Graph Technology
The document discusses the advantages of graph technology, particularly focusing on Neo4j's graph data science capabilities. It highlights various applications of graph algorithms in data analysis and machine learning, along with examples of successful implementations in banking and fintech. The text emphasizes the significance of graph structures in gaining insights and improving data-driven decision-making.
More Related Content
Similar to ntroducing to the Power of Graph Technology
ntroducing to the Power of Graph Technology
- 1. Neo4j, Inc. Allrights reserved 2021 Neo4j, Inc. All rights reserved 2021 1 Introducing to the Power of Graph Technology Kristof Neys, Graph Data Science Specialist, Field Engineering EMEA/APAC May 2022
- 2. Neo4j, Inc. Allrights reserved 2021 2 Why Graphs?
- 3. Neo4j, Inc. Allrights reserved 2021 3 Driving Intelligence into Data with Knowledge Graphs Data Graph Dynamic Context Knowledge Graph Deep Dynamic Context
- 4. Neo4j, Inc. Allrights reserved 2021 User :VISITED Website User IPLocation Website IPLocation Website Website Website :VISITED :VISITED :VISITED :USED :USED : U S E D : V I S I T E D : V I S I T E D :VISITED :SAME_AS Graphs allows you to make implicit relationships…. ….explicit Graphs….Grow!
- 5. Neo4j, Inc. Allrights reserved 2021 :SAME_AS User :VISITED Website User IPLocation Website IPLocation Website Website Website :VISITED :VISITED :VISITED :USED :USED : U S E D : V I S I T E D : V I S I T E D :VISITED User :SAM E_AS :USED :VISITED PersonId: 1 PersonId: 1 PersonId: 1 User PersonId: 2 :VISITED …and can then group similar nodes…and create a new graph from the explicit relationships… A graph grows organically - gaining insights and enriching your data Graphs….Grow!
- 6. Neo4j, Inc. Allrights reserved 2021 Not that long ago…. Deepmind stated... “We argue that combinatorial generalisation must be a top priority for AI to achieve human-like abilities, and that structured representations [i.e. Graphs] and computations are key to realizing this objective”
- 7. Neo4j, Inc. Allrights reserved 2021 Everything is a Graph...
- 8. Neo4j, Inc. Allrights reserved 2021 8 results from https://dimensions.ai, a site that tracks research papers. The search was for "graph neural network" OR "graph convolutional" OR "graph embedding" OR "graph learning" OR "graph attention" OR "graph kernel" OR "graph completion" Because I say so others say so!
- 9. Neo4j, Inc. Allrights reserved 2021 Graph Neural Networks are HOT!
- 10. Neo4j, Inc. Allrights reserved 2021 “By 2025, graph technologies will be used in 80% of data and analytics innovations...” Top 10 Trends in Data and Analytics, 11 May 2020, Rita Sallam et al.
- 11. Neo4j, Inc. Allrights reserved 2021 11 What can Neo4j Graph Data Science do?
- 12. Neo4j, Inc. Allrights reserved 2021 Neo4j’s Graph Data Science Framework Neo4j Graph Data Science Library Neo4j Database Neo4j Bloom Scalable Graph Algorithms & Analytics Workspace Native Graph Creation & Persistence Visual Graph Exploration & Prototyping
- 13. Neo4j, Inc. Allrights reserved 2021 13 Graphs & Data Science Knowledge Graphs Graph Algorithms Graph Native Machine Learning Find the patterns you’re looking for in connected data Use unsupervised machine learning techniques to identify associations, anomalies, and trends. Use embeddings to learn the features in your graph that you don’t even know are important yet. Train in-graph supervised ML models to predict links, labels, and missing data.
- 14. Neo4j, Inc. Allrights reserved 2021 14 Before we go any further…let’s quiz!
- 15. Neo4j, Inc. Allrights reserved 2021 15 Which of the colored nodes would be considered the most ‘important'?
- 16. Neo4j, Inc. Allrights reserved 2021 16 Which of the colored nodes would be considered the most ‘important'?
- 17. Neo4j, Inc. Allrights reserved 2021 17 60+ Graph Data Science Techniques in Neo4j Pathfinding & Search • Shortest Path • Single-Source Shortest Path • All Pairs Shortest Path • A* Shortest Path • Yen’s K Shortest Path • Minimum Weight Spanning Tree • K-Spanning Tree (MST) • Random Walk • Breadth & Depth First Search Centrality & Importance • Degree Centrality • Closeness Centrality • Harmonic Centrality • Betweenness Centrality & Approx. • PageRank • Personalized PageRank • ArticleRank • Eigenvector Centrality • Hyperlink Induced Topic Search (HITS) • Influence Maximization (Greedy, CELF) Community Detection • Triangle Count • Local Clustering Coefficient • Connected Components (Union Find) • Strongly Connected Components • Label Propagation • Louvain Modularity • K-1 Coloring • Modularity Optimization • Speaker Listener Label Propagation Supervised Machine Learning • Node Classification • Link Prediction … and more! Heuristic Link Prediction • Adamic Adar • Common Neighbors • Preferential Attachment • Resource Allocations • Same Community • Total Neighbors Similarity • Node Similarity • K-Nearest Neighbors (KNN) • Jaccard Similarity • Cosine Similarity • Pearson Similarity • Euclidean Distance • Approximate Nearest Neighbors (ANN) Graph Embeddings • Node2Vec • FastRP • FastRPExtended • GraphSAGE • Synthetic Graph Generation • Scale Properties • Collapse Paths • One Hot Encoding • Split Relationships • Graph Export • Pregel API (write your own algos)
- 18. Neo4j, Inc. Allrights reserved 2021 18 How can they be used? Stand Alone Solution Find significant patterns and optimal structures Use community detection and similarity scores for recommendations Machine Learning Pipeline Use the measures as features to train an ML model 1st node 2nd node Common neighbors Preferential attachment Label 1 2 4 15 1 3 4 7 12 1 5 6 1 1 0 18
- 19. Neo4j, Inc. Allrights reserved 2021 19 Our Implementations are Fast - and Getting Faster LDBC100 (LDBC Social Network Scale Factor 100) 300M+ nodes 2B+ relationships LDBC100PKP (LDBC Social Network Scale Factor 100) 500k nodes 46M+ relationships Logical Cores: 64 Memory: 512GB Storage: 600GB NVMe-SSD AWS EC2 R5D16XLarge Intel Xeon Platinum 8000 (Skylake-SP or Cascade Lake) Node Similarity 20min Betweenness Centrality 10min Node2Vec 2.8min Label Propagation 46sec Weakly Connected Components 36sec Local Clustering Coefficient 4.76min FastRP 1.33min PageRank 53sec Louvain 14.66min
- 20. Neo4j, Inc. Allrights reserved 2021 20 It’s all about Embeddings…
- 21. Neo4j, Inc. Allrights reserved 2021 Node Embedding What are node embeddings? How? The representation of nodes as low-dimensional vectors that summarize their graph position, the structure of their local graph neighborhood as well as any possible node features Encoder - Decoder Framework
- 22. Neo4j, Inc. Allrights reserved 2021 Node Embedding
- 23. Neo4j, Inc. Allrights reserved 2021 Graph Embeddings in Neo4j Node2Vec Random walk based embedding that can encode structural similarity or topological proximity. Easy to understand, interpretable parameters, plenty of examples GraphSAGE Inductive embedding that encodes properties of neighboring nodes when learning topology. Generalizes to unseen graphs, first method to incorporate properties FastRP A super fast linear algebra based approach to embeddings that can encode topology or properties. 75,000x faster than Node2Vec extended to encode properties
- 24. Neo4j, Inc. Allrights reserved 2021 24 Graph Machine Learning
- 25. Neo4j, Inc. Allrights reserved 2021 25 Node Classification - in Neo4j Load your in- memory graph with labels & features Use nodeClassification.train Specify the property you want to predict and the features for making that prediction Node classification: Predicting a node label or (categorical) property Neo4j Automates the Tricky Parts: 1. Splits data for train & test 2. Builds logistic regression models using the training data & specified parameters to predict the correct label 3. Evaluates the accuracy of the models using the test data 4. Returns the best performing model The predictive model appears in the model catalog, ready to apply to new data
- 26. Neo4j, Inc. Allrights reserved 2021 26 Link Prediction - in Neo4j Load your in- memory graph with labels & features Use linkPrediction.train Split your graph into train & test splitRelationships.mutate Link Prediction: Predicting unobserved edges or relationships that will form in the future Neo4j Automates the Tricky Parts: 1. Builds logistic regression models using the training data & specified parameters to predict the correct label 2. Evaluates the accuracy of the models using the test data 3. Returns the best performing model The predictive model appears in the model catalog, ready to apply to new data
- 27. Neo4j, Inc. Allrights reserved 2021 27 Link prediction
- 28. Neo4j, Inc. Allrights reserved 2021 28 Link Prediction
- 29. Neo4j, Inc. Allrights reserved 2021 29 Validation of Graph Technology in Banking
- 30. Neo4j, Inc. Allrights reserved 2021 30 Knowledge graphs in Credit risk analysis
- 31. Neo4j, Inc. Allrights reserved 2021 31 Stock Market sentiment
- 32. Neo4j, Inc. Allrights reserved 2021 32
- 33. Neo4j, Inc. Allrights reserved 2021 33 Neo4j working with AWS Sagemaker
- 34. Neo4j, Inc. Allrights reserved 2021 Neo4j in the AWS Ecosystem AWS Cloud Kafka Connect Plugin Connector for Apache Spark Neo4j Graph Data Science Neo4j Graph Database Neo4j Bloom Database Business Intelligence Analytics Connector for BI Amazon S3 Amazon SageMaker Amazon Managed Streaming for Apache Kafka Amazon QuickSight Amazon EMR
- 35. Neo4j, Inc. Allrights reserved 2021 Neo4j and SageMaker 1. Generate graph feature embeddings in Neo4j Graph Data Science 2. Export to S3 3. Import into SageMaker 4. Supervised Learning AWS Cloud Amazon SageMaker Amazon S3
- 36. Neo4j, Inc. Allrights reserved 2021 36 Banking Fraud: A use case
- 37. Neo4j, Inc. Allrights reserved 2021 37 Accelerated Fraud Detection in FinTech • As part of a larger business transformation initiative, they wanted to reduce cost and time associated with money transfers • False negatives in AML/Fraud create delays during investigation and ultimately unhappy customers • From established base in EU, rapid expansion into US and APAC meant compliance to new regulations and laws • Current home-grown solution was too slow (manual), expensive and didn’t scale as fast as the business growth • Traditional rules-based approach only focused on known issues while fraudsters “think ahead” The Challenge: “We had been an Insights customer for years with Synthetics but needed to better understand the real experience of our customers and how they were impacted by changes to the site. Business Analytics with Insights solved that for us.” ● International B2B payment delivery and banking services provider Founded in 2013, headquartered in Luxembourg with 200 customers ● Processing 6% of European B2C e-commerce payments in 2020 and over 250 billion Euros in payments volumes ● Delivers rapid access to direct clearing and partner banks enabling cross-border payments in 25 currencies • Required an on-going, scalable and supported solution as opposed to “throwing bodies at the problem” • Sought a forward-looking, Machine Learning (ML) solution compatible with their AWS architecture • Insights customer for 10 years • Dynatrace DEM (RUM + Synthetic) for 2 years The Requirements:
- 38. Neo4j, Inc. Allrights reserved 2021 38 Reduced false negatives & alerts = ROI in months “We had been an Insights customer for years with Synthetics but needed to better understand the real experience of our customers and how they were impacted by changes to the site. Business Analytics with Insights solved that for us.” “For AML, when you visualize all the connections on a screen, you can very easily spot important items: what used to take 3+ days to look for a connection can be found in less than 30 seconds.” Ruben Menke - Sr Data Scientist, Banking Circle • Reduced false negatives by 25% • Decreased numbers of overall alerts escalated for manual reviews by 50% • Empowered non-technical users (i.e. investigators) to gain instant insights in graph form • ML approach optimizes information from customers to see patterns and build models based on real-time data The Solution:
- 39. Neo4j, Inc. Allrights reserved 2021 Neo4j and AWS ● >40% of Neo4j customers run on AWS ● Member of Amazon Partner Network since 2013 ● APN Advanced Tier Partner ○ AWS ISV Workload Migration ○ APN Global Startup ○ ISV Accelerate ● Collaborative Joint Engineering
- 40. Neo4j, Inc. Allrights reserved 2021 On the AWS Marketplace: ● Neo4j Enterprise ○ AMI and CFT ○ BYOL ○ Graph Database ○ Graph Data Science ○ Bloom ● AuraDB Enterprise ○ DBaaS Getting Started
- 41. Neo4j, Inc. Allrights reserved 2021 Neo4j, Inc. All rights reserved 2021 41 Thank you! ● Try it yourself! ○ Neo4j-Partners GitHub ○ AWS Quick Start ● Further Reading ○ Landing Page ○ APN Partner Finder ● Contact us: ecosystem@neo4j.com Q&A