partner webcast – oracle streams analytics platform - is it time to reverse traditional analytics?
Post on 06-Jan-2017
386 Views
Preview:
TRANSCRIPT
Copyright © 2014 Oracle and/or its affiliates. All rights reserved. |
Safe Harbor Statement
The following is intended to outline our general product direction. It is intended for information purposes only, and may not be incorporated into any contract. It is not a commitment to deliver any material, code, or functionality, and should not be relied upon in making purchasing decisions. The development, release, and timing of any features or functionality described for Oracle’s products remains at the sole discretion of Oracle.
Oracle Data Integration Solutions (DIS) Oracle Stream Analytics – Is it time to reverse analytics?
Milomir Vojvodic & Alessandro Cagnetti – EMEA Data Integration Product Team
Eight Core Products
Cloud or On-Premise
Business Friendly
Extreme Performance
Spatial Awareness
Oracle Stream Analytics
DB
Web / Devices
Data Event
Data & Transaction Streams
Downstream (eg; Hadoop)
Data Event
Oracle Stream Analytics is a powerful analytic toolkit designed to work directly on data in motion – simple data correlations, complex event processing, geo-fencing, and advanced dashboards run on millions of events per second.
Innovative dual model for
Apache Spark or Coherence grid
Simple to use spatial and geo-fencing features an industry first
Includes Oracle GoldenGate for
streaming transactions
Traditional Analytics Architecture
Target DB
OGG
Source DB
Analytics Transactions
• Data set Architecture
• Data Static & Query Dynamic
• Streaming Architecture
• Query Static & Data Dynamic
7
• What it does: • Compelling, friendly and visually stunning real time streaming analytics user experience for Business users to dynamically create and implement Instant Insight solutions
• Key Features • Analyze simulated or live data feeds to determine event patterns, correlation, aggregation & filtering
• Patterns library for industry specific solutions
• Streams, References , Maps & Explorations abstracted integration , Data Discovery Canvas
• Benefits •Accelerated delivery time from months to minutes
• Hides all the challenges & complexities of underlying real time event driven analytics infrastructure
OSA Introduction
Adapter Cache Processor POJO
EPN (Event Processing Network) Elements
Channel
Input event streams
Output event streams
• Application logic contained in processor nodes
• Programmed in Java and Continuous Query Language (CQL)
Event Processing Network Of OSA
• The following streaming operators are the fundamental building blocks of streaming analytics:
• Transformation - narrow the incoming stream. A light version of traditional extract, transform, load (ETL) for streaming applications.
• Correlation - combine data from multiple sources.
• Enrichment - reference data to provide additional context.
• Time windows - snapshot of the stream over an arbitrary time period. Perform time series analysis in real time, such as running totals, weighted moving averages.
• Pattern matching - patterns that only emerge as new streaming data arrives.
• Business logic - the result of stream analysis is to inform applications with real-time context.
Blocks Of Streaming Analytics
• New Extensive Patterns Library, including Spatial, Statistical, General industry and Anomaly detection, streaming machine learning
• Streaming Expression and Business Rules Analysis
• Simplistic definition of Event Streams and References
• Actions to act on Analysis, push downstream to various event sinks, including exporting to Jdeveloper
• Visual GEOProcessing with GEOFence relationship spatial analytics
• Abstracted visual façade to interrogate live real time streaming data and perform intuitive in-memory real time business analytics
• Graphical representations of tabular streaming information
• Catalog Topology Viewer and Navigation
• An array of new streaming end point connections/targets, including Kafka
• Catalog perspectives for major industries
Main Features Of OSA
First Time Oracle platform on this report for 7 years Oracle delivers Two distinct unique pieces that are critical for the future of analytics: Stream Explorer for ingesting and interrogating data as it lands in the cloud or the enterprise; and Oracle Edge Analytics (OEA) for filtering, aggregating, and preprocessing data on embedded devices.
Report based on last years (March), first release of Stream Explorer product, with no Spark Streaming, no Kafka, and without the array of new features in the Oracle Stream Analytics product released this April
2014
Analysts
• Enhanced Patterns Library
• New Geo-spatial pattern
• Integrated Expression Builder
• Support for Business Rules in Explorations
• New streaming end point connections/targets
• New Event Stream sources and targets, such as MQTT, Apache Kafka and Twitter
• Scaling-Out with Spark Streaming
• Better Insights with Catalog Topology Viewer and Navigation
New Features Of OSA
Utilities & Oil and Gas Internet of Things Public Sector
Financial Services Transportation &
Logistics Telecommunications Manufacturing & Retail
Vertical Adaption
• Fire emergencies. Immediately isolate the fire location and define exclusion zones around the incident. In parallel, the identification of available fire resources, that are best equipped and in nearest vicinity is vital.
• Toll System - Charging a single price for every vehicle, regardless of time of day or is not an efficient model. Tolls can be computed dynamically based on congestion, accidents, and strategies for optimizing traffic.
• Airlines: Monitor all airline's operational events to eliminate flight delays, create passenger alerts , detect baggage location due to local or destination-city weather, ground crew operations, airport security, etc. Analyze event data directly aircraft, as soon as they land, and together with historical trending event data, the flight readiness can be instantly determines for the next flight.
• Air Defense - the flight readiness can be instantly determines for the next flight.
• Vehicle telematics : Reduce fuel cost alerting on element of the vehicle, Improved safety by out-of-hours usage, transgressing unscheduled locations
• Supply Chain and Logistics: Detect and report on potential delays in arrival.
Use Cases
• IT. Entire system infrastructure running at optimum performance ( online retail ). Analyze : SNMP traps indicating HW status, OS warning and error events (log “tailing”), MW events, EDN, ESB and events from the executing critical Apps
• Financial Services: Banking. Immediate Action – Payments processed more than 60 minutes without ACK or Bank Error. Ability to perform real-time risk analysis, securities trading and foreign exchange prices. Online Fraud - anticipating trends prior to successive attempts. Insurance: In conjunction with Oracle Real Time Decisions, ability to learn to detect potentially fraudulent claims.
• Retail: Proximity based marketing to provide personalized offers
• Telco: Location based offers, real-time call detail (CDR) record monitoring
• Utilities - Smart Meters “influence” individual meters to reduce home consumption levels.
• Oil&Gas- monitor oil pressure levels in real time and weather conditions and various resources nearest to the oil pipe with the correct skills to correct the problems.
• Healthcare: Medical Device Data to help save lives. Smart beds : Body
Use Cases
DEMO Oracle Stream Analytics in action
• Use separate optimized platforms for each workflow stage:
– Oracle R Enterprise (ORE) component of Oracle Advanced Analytics (OAA) for model development and building
– Oracle Stream Explorer (OSX) for streaming data scoring, stream pre/post-processing
Input event stream
Model Builder
Model Scoring
Model transfer from ORE to OSX using PMML
Oracle R Enterprise
Oracle Stream Explorer
Preprocessing, model import & scoring function creation
Training data
Prediction stream
Streaming Machine Learning Architecture
Header
Data Dictionnary
Data Transformations
Scope of Fields
Mining Schema
Targets
Outputs
Mapping of user data in suitable form for the DM model
Definitions for all fields used by the DM model (data types, ranges …)
Taxonomies Multiple Models
Model Verification
Fields usage (active/target), policies for handling missing/invalid values, outliers …
….
Common syntax for handling target categories & post-
processing
Suitable form of output fields
Version/timestamp/model development environment …
• Classes of models supported by PMML
• Association Rules
• Baseline Models
• Decision Trees
• Center- & Distribution-based Clustering
• Regression & General Regression
• k-Nearest Neighbors
• Neural Networks
• Naïve Bayes
• Sequences
• Text
• Times Series
• Support Vector Machines
PMML
Q&A
Milomir Vojvodic & Alessandro Cagnetti – EMEA Data Integration Product Team Migration Center blog: http://blogs.oracle.com/imc Migration Center email: partner.imc@beehiveonline.oracle.com
Oracle Partner Hub ISV Migration Center
Oracle.com Partner Hub
Team Info, Events/Activities Schedule, etc
Migration Center Team Blog
Webcasts, Howto, Demos, Guides, etc Youtube: OracleIMCteam
Slideshare: Oracle_IMC_team
twitter.com/OracleIMC
plus.google.com/+OracleIMC
facebook.com/OracleIMC
linkedin.com/groups/Oracle-Partner-Hub-Migration-Center-4535240
feeds.feedburner.com/oracleimc
Partner.IMC@beehiveonline.oracle.com
top related