Role of Data Integration | My Assignment Tutor

2 0 AUGUST/SEPTEMBER 2016 I DBTA S ponsored C ontentORACLERole of Data Integrationin Unlocking the Valueof Big Data SolutionsT hree years back, you would have had tofollow secret directions to the basementto find out where your organization’s bigdata lab was situated. If lucky, you mighteven have spotted a white coated datascientist. Fast forward to today, and youwould be hard pressed to find a meetingroom where senior executives are nothuddled around a white screen with bigdata written in bold red.This is not surprising. The big dataecosystem is forever changing and it playssuch an important role that it figuresin many strategic projects with some ofthe best brains in the enterprise headingthem. Compounding the problem isthat the big data technology landscapeis no longer clearly partitioned into datamanagement and data visualization. Ithas expanded to include data acquisition,data wrangling, data movement, dataprocessing, deep data analytics, datavisualization and discovery, and finallydata governance solutions.Big data integration and big datagovernance touch all these subdivisions ofdata integration.EMERGENCE OF DATAWRANGLING IN THEBIG DATA ERAThe criteria to invest in big datasolutions have seen a marked shift fromtechnology-based decisions to those basedon business use cases and benefits. Whereearlier, purchasing decisions were madebased significantly on addressing technicallimitations of current solutions, todayorganizations have started viewing big datatechnologies as a way to solve businessproblems or even transform the waythey do business. In other words, big datasystems have become another data sourceor target that serves key business initiatives.Data preparation or data wranglinghas emerged as a leading use case whichwas almost non-existent before this shift.Data preparation and data set creation hastraditionally been a bottleneck for quicklyunlocking value from business data. Datapreparation solutions use a big data-basedprocessing engine like Spark for a betteruser experience. Data wrangling serviceshave brought together traditional ETLand data discovery features by providingeasy data set creations, data enrichments,publishing and operationalizing (i.e.,automating the entire data preparationprocess) capabilities. Data curation is nolonger the sole domain of IT experts. Thisdecoupling of business dependency onIT resources and expertise has enabledfaster data analysis, enabling customersto unleash their business savvy ontounsuspecting data sets. Data wranglingtechnologies have also merged the linebetween data quality through dataenrichment; ETL, through easy dataacquisition; and data discovery, throughintelligent and rich recommendations.Leaders in data preparation solutionsuse advanced machine learningalgorithms and natural languageprocessing to glean and bubble upinsights from large and varied datasets that cannot be manually discernedbecause of the scale and complexity of thedata sets.This does not mean that the classic dataanalyst roles, those of the business analyticsand ETL experts, are becoming obsolete.In fact, their roles have evolved to enablingenterprise data management wherethey are called upon for more strategicinitiatives like moving the infrastructure tothe cloud and ensuring data for Softwareas a Service (SaaS) applications.What really differentiates the multitudeof big data projects within organizationsis not just the functional sophisticationof the projects, but the speed of deliveryof insights. While democratizing dataand enabling business users to interactdirecdy with data are irreplaceable tohelp make data-based decision makingubiquitous within the organization, it isthe actual technological and architecturaladvancements in big data technologies thatenables faster decisioning.HOW FAST IS FASTENOUGH FOR ANALYTICSSTREAMING ANALYTICSThe short answer is pretty fast.Real-time data is critical to makingany business decision more pertinent.Streaming data right into cloud datawarehouses, which are in turn pluggedwith data visualization solutionsdownstream, ensures that the data beingused for business analytics is the latestdata. Successful organizations invest innot just data visualization tools, which aretypically the tip of the data managementiceberg. They understand and implementa responsive and governed network ofdata delivery pipeline to capture, filter,aggregate and correlate data before itreaches the analytics platform.Streaming data first captures data rightat the source of the data creation. However,most of the events that generate the datahappen on business-critical applicationsand systems. Real-time data integrationtechnologies should balance the need tocapture data in real-time without slowingthe performance of the systems.Combining big data with streaming datacapabilities presents infinite possibilities.Data from transactional systems that iscaptured in the organization’s databasecombined with user generated data inreal-time finds many applications in thereal world. Businesses with data-drivenmarketing initiatives can improve customerexperience by generating customizedpromotional offers based on varioushistoric factors such as purchase historycombined with real-time data such aslocation, and data from their social mediafootprint. Today, real-time data integrationtechnologies are used mainly to deliver datato the analytics data warehouse or datamarts. There is an opportunity to embedORACI_€Eanalytics in the data streams that deliver thedata to accelerate insights and action.Streaming analytics that includeevent stream processing shortens thetime from data creation to decisiondrastically. Stream analytics empowersa business audience in any industrythat is looking to create solutions thatembrace real-time, instant insight acrossdata delivery infrastructures. A gooddeep data storage provides the best ofboth worlds. Lambda architectures, assuch combined streaming and deep datastorage architectures are called, providea single platform that enables enterprisesto perform both real-time streaminganalytics, and refine the analytics withinsights from mining for richer andmore complex data recommendationsfrom the deep data storage reservoir.NoETL Engine100% Native DataTransformationNon-invasive CDC,Realtime streamingdata deliveryPrepare, Secure,Enrich and PublishUnstructured DataCatalog, Trace andView Models across .the EnterpriseFederate DataAcross DBs, Servicesand ApplicationsO racle offers a fu ll set o f p ro d u c ts fo r cloud, b ig data a n d on-pre m ise data in te g ra tio n requirem entsstream analytics solution is designedto handle large data volumes withsubsecond latency, while also providinga business-friendly, easy-to-use interface.They are designed with drag-and-dropinterfaces that help model data streamsreplicate business models and behaviors.Streaming analytics finds great use inscenarios that rely on very low latencybusiness decisioning. Some examplesinclude fraud detection in the financialindustry to automate stock tradingbased on market movement, monitoringthe vital signs of a patient and settingpreventative triggers in health care, anddetecting security issues and fraud intransportation industries by findinganomalous patterns as they happen toinitiate immediate investigation.Stream analytics combined withBecause big data technologies underpinthe data storage, the cost-to-benefit ratiois extremely appealing for organizationslooking to make a difference with theirbig data investments.THE ORACLE ADVANTAGEOracle provides a wide range of productsthat help with all the moving parts ofbuilding a differentiated and forwardlooking big data integration, managementand analytics platform. As part of the dataintegration portfolio, Oracle GoldenGateand Oracle GoldenGate Cloud Serviceensure real-time data capture andstreaming from heterogeneous businesscritical transactional systems with minimalimpact to the performance of the sourcesystems. Oracle GoldenGate providesthe most secure and reliable big datadelivery solution between the cloud andon-premise systems and applications.Oracle’s Big Data Preparation CloudService is a next-generation data wranglingservice that helps business users unlockdata quickly from complex business data.Oracle Big Data Preparation Cloud Serviceis built on Apache Spark, combines naturallanguage processing and machine learningand bridges the line of business—IT dividewhen extracting insights from the dataand operationalizing them into enterprisedata integration flows. Meanwhile, theOracle Stream Analytics platform providesa compelling combination of an easyto-use visual facade to rapidly create anddynamically change real-time event streamprocessing applications, together witha comprehensive run-time platform tomanage and execute these solutions. Thistool is business user-friendly and solves thebusiness dilemmas by completely hidingand abstracting the underlying technologyplatform. Oracle Data Integrator bringstogether big data platform portability, theability to switch between multiple big dataplatforms seamlessly, and powerful datatransformation capabilities to enterprisebig data development teams. Oracle DataIntegrator provides a unified and commoninterface to build data transformationsirrespective of the underlying big datatechnology. This ensures the dataintegration developers can utilize the latestbig data platform and language withouthaving to compromise on productivityand with minimal disruption. Oracle’sentire integration platform is governed andaudited by Oracle Metadata Management.Oracle big data integration offerings areflexible, robust and complementary tomaximize big data investments and unlockvalue from these investments now and inthe future. ■ /goto/dataintegrationCopyright of Database Trends & Applications is the property of Information Today Inc. andits content may not be copied or emailed to multiple sites or posted to a listserv without thecopyright holder’s express written permission. However, users may print, download, or emailarticles for individual use.


Leave a Reply

Your email address will not be published. Required fields are marked *