Data Quality
Cognity is a Partner of Trillium Software for Total Data Quality solutions.
Data Quality solutions are critical components in successful data integration, Customer Relationship Management (CRM), Business Intelligenece (BI), Analytics, Data Warehouse (DW), Business Performance Management (BPM), Enterprise Resource Planning (ERP), and Supply Chain Management (SCM) projects. We can provide you with a comprehensive solution of software and services for greater global information integrity, wherever and however data is captured, collected, stored and manipulated within the enterprise.
Our solutions fully support the complete lifecycle of data management that includes data assessment and profiling, data standardisation, information enrichment, data linking and perpetual data quality monitoring. In each phase of this lifecycle, raw, incosistent and chaotic data become more valuable and usable information that business users can trust and managers can use to optimise business performance.
Designed for diverse enterprise environments and technology architectures, the scalable and modular components of the solution for global data quality and data profiling can be ported across diverse systems, integrated with enterprise applications, iteratively tuned for consistent results and globally enabled for business without borders.
Maintaining high data quality in this environment requires the ability to understand and improve data quality during integration and migration projects, within real-time transactions and for third-party data feeds. We provide a Universal Architecture which answers these challenges with a scalable, flexible framework that supports the integration of data quality processes into any system, at any time, anywhere in the world. From tactical projects to strategic practices, the Universal Architecture increases integration efficiency, lowers development costs, and provides faster return on investment (ROI) from data quality initiatives through:
- Modular software design - It comprises of self-contained, linked modules that you can configure and arrange to meet unique business needs. The ability to call any Total Data Quality process-Discovery, Standardisation, Cleansing, Enrichment, or Linking- individually from an application or from another data quality process lets you integrate only the data quality processes you need, when you need them, in the order you want them
- Universal connectivity components - It includes an Integration Layer that streamlines your ability to incorporate data quality processes into any data flow. The components offered are Web Services (WS), connectors to ERP and CRM Systems, ETL Tools, EJBs, APIs for calling the DQ processes from C/C++/C#, VB/VB.net, Java, RPG and COBOL
- Architecture-neutral core technology - It protects your ability to integrate data quality processes across even the most complex IT environments-in both batch and real-time business processes
- Portable, reusable resources and tunable processes - The non-proprietary text format of the resource files, such as business rules, directories, and parameters, facilitates near-instantaneous replication and portability across practically any platform or system. In addition, multiple implementations can refer to a single, centralized set of resources. Both methods let you leverage efforts from one implementation across new projects and the entire enterprise, dramatically reducing costs in multiple implementations and allowing you to easily create, propagate, and maintain an enterprise data quality standard
- Expandable, global support - Wherever your market evolves, we can support your business. Offering geographic data validation and cleansing for every country in the world and the most robust available support for more than 30 major global markets, we let you easily add country-specific data quality processes to meet new data management needs. We offer Unicode and double-byte character support, along with support for more than 30 common code pages; automatic recognition and routing of commingled country data
Methodology
Wherever and however your organisation uses data, our Total Data Quality Methodology offers a guideline for confidently implementing, integrating, and expanding data quality coverage across your organisation. Business-born and field-tested, the Total Data Quality Methodology gives you the tools and technique to bring data quality up to your organisational standard, any time data is captured, collected, stored, or manipulated across the enterprise.
Within the methodology, core data quality processes build upon each other to create an end-to-end solution for greater data accuracy, relevancy and reliability in diverse systems across the organisation:
Data Discovery and Profiling
Do you know what you don't know? Unrecognized data conditions-missing values, misaligned relationships, and broken business rules-are leading factors in stalled and derailed integration and business projects. Before you can improve your data, you must understand it-both what it contains and what it doesn't:
- Frequency of occurrences in a field, including blanks, zeros
- Shapes of data in a field, such as xxx-xxxxxxx for a phone number
- Distribution of business addresses vs residential addresses
- Data values, statistics, frequencies, and ranges
- Mismatches and inconsistencies between metadata and actual data content
Data Standardisation
Once you determine your data character, data standardisation brings consistency to your data and aligns data shapes, content, and formats to your organisational standard:
- Customer name and address data: Standardisation of names (e.g., Bob, Robert, Rob), business names (e.g., ATT, AT&T, AT and T) and address formats. Any misplaced or misspelled elements can be corrected
- Business data: Normalisation of product information, part numbers, tax ID numbers, telephone numbers, and other generic information
- Global data: Country- and culture-specific data standardisation
Standardisation creates an organised data foundation for more accurate linking, more relevant enrichment, and more rapid data integration.
Information Enrichment
By completing, augmenting, and correcting existing records, enrichment can increase data value in many ways; for example, by:
- Adding business value to data without the overhead associated with its capture.
Completing fractional records
- Enhancing records with new data from third-party and other operational sources
- Correcting data that appears right, but isn't, such as misspelled city names
- Building valid addresses from fragmentary data elements
- Providing very complex data such as latitude and longitude that can't be entered into a system by conventional means
Linking
Linking recognizes relationships within records based on commonalities in data content to create multi-dimensional, unified views of business entities. Sophisticated linking can recognize several levels of customer and business relationships based on your specified criteria. Here are just a few common linking scenarios:
- Increased call-center efficiency through real-time customer recognition and record completion
- Identification of customer affinity groups by linking records based on marketing codes
- Stronger profiling and target marketing by linking customer records based on demographic information provided through enrichment
- Better understand of total inventory or sales for individual products across distributed points of sale (POS) via real-time order entry linking
- Elimination of redundant marketing and recognition of cross-sales opportunities through household identification
- Recognition of broken business rules by identifying where objects that should always be joined aren't (e.g., product sales and customer accounts)
Integration
Because data quality is not a standalone process, the ability to integrate data discovery and improvement seamlessly into existing business processes is critical. The system supports integration through a universal architecture and purpose-built Data Quality Connectors for leading enterprise applications, including Siebel eBusiness, SAP NetWeaver, Oracle E-Business, Informatica, Microsoft DTS, and IBM DB2.
|
