Analytics and Data Summit 2020

Graphs enable intuitive detection of complex relationships and navigation of connections for new insights into your data. Powerful algorithms for ranking, centrality, community detection and path finding are available to analyze data with graphs, supporting a variety of use cases such as fraud detection, recommendation engines, dependencies in manufacturing, and more. In this session, learn how to model a graph, load a graph, insert nodes and edges using graph APIs, query a graph using PGQL (Property Graph Query Language), analyze a graph using 50+ pre-built algorithms available in the Java API, and visualize a graph using a native web-based tool. We will walk through an example that provides an end-to-end view of how to build an application that uses graph analytics.

This session provides a preview of planned future functionality. Low-code, automated services for working with graphs in Oracle Autonomous Data Warehouse Cloud Service will simplify the creation and management of graphs for powerful analytics. Learn how to use automated tools to create a graph model from database tables and perform a variety of graph analytics. In the Cloud service you can navigate relationships, identify patterns, perform connectivity-based queries, and discover anomalies using pre-built analytics algorithms. You can customize visualization of graph analytics results: as a graph, as a tag cloud, as a chart, and more. We will use a digital marketing example to walk through a complete workflow.

Graphs are one of the leading trends, Gartner identify graph analytics as an emerging technology people should experiment with over the next year.Many sessions cover the property graphs topic and what they can do and how graph analytics is powerful. Some cover the technical aspects of what product has support for property graphs. What is still missing? A simple thing as 'how to start?'.How does one create a property graph? What tools are needed? How to integrate property graphs with the existing tools and data flows? (database, files, Kafka streams etc.) This hands-on training will cover that gap between theory and graph analytics: how to get started, how to integrate things. From nothing to a graph first, to cover the basis. Moving to how to load a property graph in an Oracle Database and how to use it for graph analysis later.There will also be an introduction of the Java client and the REST interface, to provide all the tools for attendees to design and build graphs.

To gain experience and grow skills with the Spatial & Graph library that is now available in Oracle's Autonomous Data Warehouse (ADW), we have built a network graph on top of railroad data that the government of India makes available at https://data.gov.in/resources/indian-railways-time-table-trains-available-reservation-01112017. We suspect that our experiences will be of interest to others desiring to build a large ADW graph for other use-cases in other industries, and the following provides a progress report with lessons learnt.

The movements of the many trains that travel across this railroad network also trace a graph whose nodes are train stations and whose edges are the connecting railroad segments. The transactional input data details the motions of roughly 10K distinct trains as they execute about 200K trips along 20K distinct rail segments that connect 10K railroad stations, all in a single day in India. So our first step is to compose an Oracle Machine Learning (OML) notebook that distills that transactional input data down to two aggregate ADW tables: one table listing the 10K nodes and their summed properties (total number of trains visiting each station & number of edges radiating from each station), and an edges table containing 20K records detailing mean train speed and distance between adjacent stations. Building the graph in ADW is then a blissfully simple SQL one-liner.

We then compose another OML notebook to ask the usual sorts of questions of this graph, such as: what is the shortest path (time-wise and distance-wise) connecting two arbitrary stations. The resulting path is then decorated with all adjacent stations that are one or two train-rides away, with those results then exposed to Oracle Analytics Cloud (OAC) as derived ADW tables. Note that the graph of the entire India railroad network is too large to visualize in OAC (the 10K edges and nodes overwhelms an OAC canvas), so we used a variety of methods to filter this graph in a way that extracts this railroad network's main artery containing the busiest nodes and edges.

This effort's final task will be to sprinkle this ADW graph with numerous virtual traveling agents, and to train a machine-learning algorithm to direct the movements of these agents across this graph in a way that maximizes the virtual rewards gathered by these agents while minimizing travel time and expense. Progress achieved on this final task will be reported at conference time, where we will also show how our solution to this Pacman-like problem is relevant to ADW users in sales and finance.

The primary value that Graph data model brings to Healthcare services is to aggregate information from various heterogeneous data sources (e.g., Electronic Medical Records from Medical Data Centers or publicly-available Disease-Symptoms relations) whereas Machine Learning provides the ability to process such huge volume of medical data and learn meaningful underlying patterns or latent relationships. In this talk, we will explain how Oracle Labs and Oracle HS-GBU employ a combination of graph and machine learning techniques to develop efficient Healthcare Services. These healthcare services enable multiple crucial medical objectives like (a) matching Rare Patients with Clinical Trials (considerably important for Cancer treatments), as well as (b) searching for Similar Patients from a Medical Database based on previous medical records of the patients.

Graph-powered Cyber-security intelligence has recently drawn a lot of attention with the emergence of the Cloud era. Cloud applications generate huge volume of logs which are essential for threat detection and investigation. While such massive volume of data enables to accurately detect multiple threats, the data heterogeneity introduced by multiple data sources makes it challenging to analyze data in the tabular data model. Graph data model bridges the limitation by connecting heterogeneous entities via relationships. In this talk, we present how graph technology enables us to build a smarter and deeper cyber-security intelligence. Specifically, we demonstrate how Oracle Labs PGX and Data Studio enable (a) SaaS Cloud Security team to do in-depth analysis of security alerts with visualization and threat hunting from multiple heterogeneous logs, and (b) ODC Moat team to develop an effective Invalid Traffic Detection model by leveraging features extracted from underlying graph.

With the growing need for volumes of data required by ML and Knowledge Bases, copying/duplicating potentially Petabytes of data is a real problem. Working with data 'in situ' is fast becoming the only viable pattern for enterprises. Additionally, heterogenous data silos are a given for big enterprises and aren't going away. Finally, incorporating 3rd party taxonomies/ontologies/datasets for enrichment are yet another example of datasources that need to be incorporated and orchestrated. This suggests the need for a hybrid approach for Knowledge Management. With careful architecture, this is where Semantic Knowledge Graphs can really shine, and in particular Oracle's implementation of Semantic Graph technology. Further, moving to the cloud offers real long-term advantages.

The National Statistics Center of Japan (NSTAC) is responsible for managing and publishing a wide variety of official statistics data in Japan, serving national and local government agencies, and the wider public. NSTAC developed the Linked Open Data (LOD) platform in 2016 with Oracle RDF Graph to support search and discovery of major statistics data such as the population census.With greater public awareness, usage of LOD increased, and published RDF triples increased to 2 billion. This created a need to improve SPARQL performance and publish a wide range of statistics in RDF format. As a solution for high volume and performance, NSTAC migrated the platform to Oracle Cloud, including Database Cloud Service with the Database In-Memory option.This improved SPARQL performance improved by 60x, and significantly expanded processing capacity.Sample applications were also developed for users unfamiliar with SPARQL. In this session, we will describe the use case and technologies.

Property graphs enjoy an ever-increasing popularity in applications that deal with connected components of all varieties; e.g., social networks, fraud detection, etc. Today there are a number of stand-alone languages used with Property Graphs including Apache Tinkerpop Gremlin, Neo4j Cypher, Tigergraph GSQL and Oracle PGQL as well as a proposal for another stand-alone language under consideration by ISO, GQL. It is well known that SQL is the most widely used database language and is the language of choice for database application developers. To support graph creation and querying, the ISO SQL 202x standard will include new extensions for graph databases, SQL/PGQ. We will show how SQL/PGQ will integrate with SQL to allow users to define property graphs based on data stored in existing tables and to query those graphs. The query part will use a combination of familiar SQL as well as a new built-in operator, which significantly simplifies typical graph queries compared to existing SQL.