Business Data Lake Conceptual Framework
Relevant Business Scenarios for the BDL
Download 493.56 Kb.
|
- Navigate this page:
- Enterprise Data Warehouse off-load
- Discovery Platform
- Big Data Apps
- Data-Driven Enterprise
- Data-Driven Ecosystem
- The concepts of the Business Data Lake
Relevant Business Scenarios for the BDLThis chapter gives an overview the main business scenarios (defined as a set of business use cases) that are enabled by Business Data Lake implementations. Their order – although not mandatory – shows a possible underlying maturity progression. Enterprise Data Warehouse off-loadThis first step is particularly relevant for enterprises that already have many large Enterprise Data Warehouses (EDWs) and experience difficulties (for instance the projected costs of additional storage) to deal with increasing data storage needs. It’s also relevant for enterprises for which identifying and setting priorities to “Big Data” business uses cases is an issue. It’s based on the assumptions (often facts for many enterprises)
that some EDWs are overpopulated by data that could be archived or data that are duplicated. It’s the case when new applications systematically create new data marts. that some EDWs are over-engineered “gold-plated” solutions for the data they host. The idea of this scenario is to create the opportunity to build a Business Data Lake by decommissioning some of these kinds of EDWs that finally do not bring enough business value (and considering their high cost). The more the storage needs are growing fast, the easier it is to initiate this scenario that will help accommodating massive data growth with existing EDW investments. Discovery PlatformDiscovery Platforms combines data storage and simple, “easy-access” distillation steps. The idea is to demonstrate – through actual selected business use cases managed in an experimental way – the benefits of the BDL. A discovery platform starts small with proof-of-concepts for business cases that are mostly aiming at increasing operational efficiency and improving existing services. It can also start with existing data that are already stored or collected by the Enterprise. It mainly leverages the fact that BDL can store and process large sets of unstructured and structured data. It also starts with “basic” (meaning proven) machine-learning algorithms (linear and logistic regression for instance). These basic analytics capabilities must be executed by data scientists who actually can state whether the analyses are legitimate or not from a statistical and mathematical point of view. Multiple flavors and instances of discovery platforms can be experimented in the same enterprise. Big Data AppsThis scenario goes one step further compared to the Discovery Platform. Analytics are designed and tuned to be integrated into analytic-led downstream applications. The goal here is generally to significantly increase top line revenues, possibly for multiple lines of business. The BDL for building Big Data Apps becomes a real enterprise asset. It’s generally best to have a single platform serving multiple lines of business. This platform should also be built with the contribution of the IT department for the design as well as for the operations. Big Data Apps generally also integrate additional innovation solutions (social, mobile). Thus they enter the field of the Open Platform 3.0 concept. Data-Driven EnterpriseA Data-Driven Enterprise is characterized by generalized Big Data Apps that create new services and/or open new revenue streams. The BDL is then considered as a tool supporting the Digital Transformation of the Enterprise Business. It’s leveraged by an internal community of contributors who continuously brings new data and develops new analytics capabilities. At the Data-Driven Ecosystem stage the same BDL is serving multiple organizations, for instance suppliers and third-party data or analytics contributors. Compared to a “single” Data-Driven Enterprise, the community of contributors is enlarged outside of your organization. Ultimately it can be an open community. This requires security and governance topics to be specifically addressed. When the ecosystem has a naturally identified leader, that leader can build the BDL on its own and open it later. When the ecosystem is created to enable multiple “equivalent” organizations to collaborate, the BDL should be seen as an outside-in platform that could be hosted and operated by a third-party organization providing the BDL “as-a-service” to the ecosystem members. The concepts of the Business Data LakeData-related concepts 
Directory: bigdata bigdata -> Use Cases from nbd(nist big Data) Requirements wg bigdata -> Use Cases from nbd(nist big Data) Requirements wg 0 bigdata -> Nist big Data Public Working Group (nbd-pwg) nbd-pwd-2015/6a,DW. abbreviated rr (M0444) Source: nbd-pwg status: Draft Title: Big Data Use Case #6 Implementation, using nbdra author: Afzal Godil bigdata -> Nist special Publication 1500-4 draft: nist big Data Interoperability Framework: Volume 4, Security and Privacy Download 493.56 Kb. Share with your friends:
|