What Is GDS and Neo4j’s GDS Library
•
6 likes•1,306 views
Speakers: Alicia Frame, Senior Data Scientist and Amy Hodler, Director Graph Analytics and AI Programs, Neo4j
Recommended
Recommended
More Related Content
What's hot
What's hot (20)
Similar to What Is GDS and Neo4j’s GDS Library
Similar to What Is GDS and Neo4j’s GDS Library (20)
More from Neo4j
More from Neo4j (20)
Recently uploaded
Recently uploaded (20)
What Is GDS and Neo4j’s GDS Library
- 1. Neo4j for Graph Data Science™ Alicia Frame @AliciaFrame1 Lead Product Manager for Graph Data Science Amy E. Hodler @AmyHodler Director, Product Marketing & Programs for Graph Data Science
- 2. 2 • Graph Data Science (GDS) • Neo4j for GDS and the GDS Library • DEMO! • Questions #GraphDataScience #Neo4j Alicia Frame Lead Product Manager Graph Data Science Amy E. Hodler Director, Product Marketing & Programs Graph Data Science
- 3. It’s Not What You Know
- 4. It’s Who You Know And Where They Are
- 5. Whose pay will increase the most?
- 6. Photo by Helena Lopes on Unsplash Network Structure is highly predictive of pay and promotions • People Near Structural Holes • Organizational Misfits “Organizational Misfits and the Origins of Brokerage in Intrafirm Networks” A. Kleinbaum “Structural Holes and Good Ideas” R. Burt
- 7. Relationships and Network Structure Strongest Predictors of Behavior & Complex Outcomes “Research into networks reveal that, surprisingly, the most connected people inside a tight group within a single industry are less valuable than the people who span the gaps ...” 7 “…jumping from ladder to ladder is a more effective strategy, and that lateral or even downward moves across an organization are more promising in the longer run . . .”
- 9. Which is why graph data science is so powerful 9
- 10. “Data science is an interdisciplinary field that uses scientific methods, processes, algorithms and systems to extract knowledge and insights from structured and unstructured data.” - Wikipedia 10 What is data science? Data scientists use data to answer questions.
- 11. Graph Data Science is a science-driven approach to gain knowledge from the relationships and structures in data, typically to power predictions. 11 What is Graph data science? Data scientists use relationships to answer questions.
- 12. Query (e.g. Cypher/Python) Real-time, local decisioning and pattern matching Graph Algorithms Global analysis and iterations You know what you’re looking for and making a decision You’re learning the overall structure of a network, updating data, and predicting Local Patterns Global Computation
- 13. Relationships and network structures are highly predictive and underutilized – and already in your data. Graph are a natural way to store and use this predictive information, but different than what you’re doing today. How do you continually put more accurate, predictive models into production quickly?
- 14. Predictive Maintenance Churn Prediction Fraud Detection Life Sciences Personalized Recommendations Cybersecurity Disambiguation & Segmentation Search & Master Data Mgmt. Graph Data Science Applications Just a few examples…
- 15. 15 • 27 Million warranty & service documents parsed for text to knowledge graph • Graph is context for AI to learn “prime examples” and anticipate maintenance • Improves satisfaction and equipment lifespan • Connecting 50 research databases, 100k’s of Excel workbooks, 30 bio-sample databases • Bytes 4 Diabetes Award for use of a knowledge graph, graph analytics, and AI • Customized views for flexible research angles • Almost 70% of CC fraud was missed • ~1B Nodes and Relationships to analyse • Graph analytics with queries & algorithms help find $ millions of fraud in 1st year Improving Analytics, ML & AI for Enterprises Caterpillar’s AI Supply Chain & Maintenance German Center for Diabetes Research (DZD) Financial Fraud Detection & Recovery Top 10 Bank
- 16. Evolution of Graph Data Science Decision Support Graph Based Predictions Predictions within a Graphs 16 Graph Feature Engineering Graph Embeddings Graph Native Learning Knowledge Graphs Graph Analytics 16
- 17. Evolution of Graph Data Science Graph Feature Engineering Graph Embeddings Graph Networks 17 Graph AnalyticsKnowledge Graphs Graph search and queries Support domain experts
- 18. Deceptively Simple Queries Collaborative filtering: users who bought X, also bought Y (open-ended pattern matching) What items make you more likely to buy additional items in subsequent transactions? Traverse hierarchies - what items are similar 4+ hops out? Difficult for RDMS systems Knowledge Graph Queries e.g. Retail Recommendation 18
- 19. Evolution of Graph Data Science Graph Feature Engineering Graph Embeddings Graph Networks 19 Knowledge Graphs Graph Analytics Graph queries & algorithms for offline analysis Understanding Structures
- 20. • Degree Centrality • Closeness Centrality • CC Variations: Harmonic, Dangalchev, Wasserman & Faust • Betweenness Centrality • Approximate Betweenness Centrality • PageRank • Personalized PageRank • ArticleRank • Eigenvector Centrality • Triangle Count • Clustering Coefficients • Connected Components (Union Find) • Strongly Connected Components • Label Propagation • Louvain Modularity • Balanced Triad (identification) Graph Algorithms & Functions in Neo4j • 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 • Degree Centrality • Closeness Centrality • CC Variations: Harmonic, Dangalchev, Wasserman & Faust • Betweenness Centrality & Approximate • PageRank • Personalized PageRank • ArticleRank • Eigenvector Centrality • Triangle Count • Clustering Coefficients • Connected Components (Union Find) • Strongly Connected Components • Label Propagation • Louvain Modularity • K-1 Coloring • Modularity Optimization • Euclidean Distance • Cosine Similarity • Node Similarity (Jaccard) • Overlap Similarity • Pearson Similarity • Approximate KNN Pathfinding & Search Centrality / Importance Community Detection Similarity Link Prediction • Adamic Adar • Common Neighbors • Preferential Attachment • Resource Allocations • Same Community • Total Neighbors ...and also Auxiliary Functions: • Random graph generation • One hot encoding • Distributions & metrics 45
- 21. Graph Algorithms e.g. Retail Recommendations Graph algorithms enable reasoning about network structure Louvain to identify customer segmentation based on topology PageRank to measure transaction volumes Connected components identify unique users Jaccard to measure purchasing similarity 21
- 22. Evolution of Graph Data Science Graph Embeddings Graph Networks 22 Knowledge Graphs Graph Analytics Graph Feature Engineering Graph algorithms & queries for machine learning Improve Prediction Accuracy
- 23. Graph Feature Engineering Feature Engineering is how we combine and process the data to create new, more meaningful features, such as clustering or connectivity metrics. 23
- 24. Evolution of Graph Data Science Decision Support Graph Based Predictions Predictions within a Graphs 24 Graph Feature Engineering Graph Embeddings Graph Native Learning Knowledge Graphs Graph Analytics FUTURE 24
- 25. for Enterprise Graph Data Science Neo4j Graph Data Science Library Scalable Graph Algorithms & Analytics Workspace Native Graph Creation & Persistence Neo4j Database Visual Graph Exploration & Prototyping Neo4j Bloom Practical Integrated Intuitive
- 26. A graph catalog that creates an efficient analytics workspace • Reshape transactional database into an in memory analytics graph, and manage these operations • Optimized for analytics with global traversals and aggregation Algorithms • Run on the loaded graph to compute metrics about the topology and connectivity • Highly parallelized and scale to billions of nodes 26 What is the GDS Library?
- 27. • Neo4j automates data transformations • Fast iterations & layering • Production ready features, parallelization & enterprise support Neo4j for GDS enterprise-grade features and scale A graph-specific analytics workspace that’s mutable – integrated with a native-graph database Mutable In-Memory Workspace Computational Graph Native Graph Store
- 28. Answer previously intractable questions with the data you already have • Deep Path Analytics & Structural Pattern Matching • Community & Neighbors Detection • Influencer and Risk Identification • Disambiguation • Link and Behavior Prediction Massive scale to 10’s billions of nodes with optimized algorithms Increase your predictive accuracy with Neo4j GDS Algorithms Take advantage of hardened, validated graph algorithms that enable reasoning about network structure.
- 29. Find Value Faster with Neo4j’s practical Graph Data Science framework Drastically simplified and standardized API that enables custom, flexible configurations Documentation, training, and examples so getting started is simple Explore graphs and algorithm results visually with Bloom Share insights across teams for better collaboration Friendly data science experience with logical guardrails like memory mgmt. Reshapping, node & relationship aggregation / deduplication and multipartite algos
- 30. 30 Simplify Your Data Science Experience Dozens of libraries, hundreds of algos & no docs! How do we shape data into a graph in the first place? We’ve picked a library...good luck learning the syntax WTF? We have to build the entire ETL pipeline for this? Are the results right? How do we get into production? Data Modeling Which Algorithms? Learn Syntax What Now? Reshape
- 31. 31 Simplify Your Data Science Experience Dozens of libraries, hundreds of algos & no docs! With Neo4j it’s already a graph We’ve picked a library...good luck learning the syntax WTF? We have to build the entire ETL pipeline for this? Are the results right? How do we get into production? Data Modeling Which Algorithms? Learn Syntax What Now? Reshape We have validated algos, clear docs, and tutorials Neo4j syntax is standardized and simplified We seamlessly reshape your data with 1 command Easily write results to Neo4j & move straight into production
- 32. 32 DEMO
- 33. • Real data from online retailer in the UK with all-occasion items • 28K nodes and 1.1M relationships • “Data mining for the online retail industry” by D. Chen et al • Neo4j 4.0 and Graph Data Science Library 1.2 33 Retail Data
- 34. 1. Customer segmentation a. Break down the graph into customers with similar buying patterns i. Similarities, Community Detection, Mutate & Export Graph 2. Item recommendations a. What item should we recommend to different customer segments? i. Co-Purchase Similarity, Centarlities, Mutate & Export Graph 3. Explore and answer our business questions a. Recommendation for a person that buys something specific? b. What to promote to drive sales in a category? 34 Retail Demo
- 35. Thank You & Questions 35 - O’Reilly Book on Graph Algorithms neo4j.com/graph-algorithms-book/ - GDS website, whitepapers, links neo4j.com/use-cases/graph-data-s cience-artificial-intelligence/ - Data for Retail example github.com /AliciaFrame/GDS_Retail_Demo - GDS Sandbox sandbox.neo4j.com/ ?usecase=graph-data-science Resources Alicia Frame Alicia.Frame@Neo4j.com @AliciaFrame1 Amy E. Hodler Amy.Hodler@Neo4j.com @AmyHodler