Products
Acuate Integrity
Acuate Merge
Acuate Dedupliction
Related Link
Services
Dashboards and Reporting
Data Quality Methodology
Clients
|
Acuate Integrity
The most complete, flexible and easy-to-use data quality management system for business improvement
Acuate Integrity is a Microsoft Windows based solution connecting to all types of data sources from SQL Server, Oracle, DB2, Microsoft Access and Text files. The solution incorporate a comprehensive rules library, advanced matching and deduplication algorithms, custom rules grammar, links to software components, parsing and matching of international addresses, interactive data entry for exceptions, metrics management and reporting.
Creating a Data Quality Standard
Integrity has two critical concepts that allow the creation of a data quality standard. Rules can be defined to process information and transform data to a desired state. These rules can be used many times to reinforce your data quality standards and are fully reusable. Domains enable individual pieces of information to be given certain characteristics. These include format information (looking at the structure of information) and lookup values (referencing a pre-defined list of values or matching against an external reference file.) Rules and domains create a standard that can apply to all your data.
Data Connectivity
Connectivity is provided through ODBC and OleDB technologies. The data connectivity layer is abstracted to provide data independence from sources enabling matching and integration from multiple data sources.
Customer Rules Grammar
Not all rules can be specified in advance. Integrity’s Customer Rule Grammar defines bespoke rules for your business using a GUI or a text editor. The rule is treated like any other and can be reused.
For example:
if orders.discountpercentage > 10 then
orders.authorityrequired = true
else
orders.authorityrequired = false
endif
Software Component Integration
Where a standard rules or a custom rule through the grammar is not sufficient to process data, then a .NET component can be integrated as a rule. This will be treated like any other rule and integrate seamlessly into Integrity. The component can be written in C#, VB, C++ or any other.NET language that supports managed code. The component must comply with the Acuate Component Rule Interface Specification.
Matching and Deduplication
A common problem in data quality is the existence of duplicate data. This can be a duplicated record within a table of information or duplicate data across multiple sources. Integrity is designed with advanced matching and deduplication algorithms to maximise the probability of finding duplicates.
When faced with large amounts of data, matching and dedupication processes have a trade off between time and quality. When matching records across two data sources, the algorithms are quadratic in performance (i.e. increase at the rate of the number of records squared), hence the greater the amounts of data the slower the process will be. However, by applying business constraints on which records you choose to match, there can be a significant improvement in performance (over 1000’s of percent) with an acceptable drop in the number of matched records found (e.g. 1-2%). Integrity allows you to manage this trade off to maximise results.
International Addressing
The Integrity addressing module matches addresses to local Postal Address Files, ensuring the best possible matches are found.
Rules Library
A set of comprehensive rules are supplied with the Integrity product. Each rule focuses on a unique aspect of data quality, for example, a rule may fix the domain part of an e-mail address, or check the gender of a title field against the forename, or deduplicate records according to a specific matching criteria, etc.
These rules have been created by working with many types of data and with many organisations. Our aim is to provide a usable set of rules that solve today’s data quality problems.
We are committed to expanding these rules on a regular basis. We derive a vast amount of information in improving data and it is our mission to pass these rules onto our customers.
Rules fall into five categories:
- Real-World Rules – these perform the basic data manipulation required to understand data, for example, checking that an e-mail has the right format, or checking for typing errors in fields, etc.
- Cross-Reference Rules – these check two or more fields with each other, for example, that an age field is consistent with a data of birth field, or a first name matches the gender of a title, etc.
- Reference Rules – these rules match to external reference files, for example, Royal Mail’s Postal Address File or Experian’s National Business Database, etc.
- Deduplication Rules – these perform matching and deduplication functions across one or multiple data sources, for example, deduplicating product codes across two sources, or deduplicating company information with a single source, etc. and
- Customer Rules – these use the Custom Rules Grammar and the integration of software components as described above.
Data Measurement
Integrity provides a flexible mechanism to measure the data quality over times. Every time a rule is executed, the improvements can be measured and targeted. If you were improving you weekly, then you can plot a graph to see how well you data quality has improved.
Lets say you wished to measure the quality of your account records. You decided to look at five aspects of your customer information: (i) whether you have a business names matched to Experian’s National Business Database, (ii) whether the address is matched to Royal Mails Postal Address File, (iii) a valid e-mail exists, (iv) a correctly formatted telephone exists, and (v) the record is unique. A metric can be created for each element and a new weighted metric (say, 50%, 30%, 10%, 5% and 5% of the respective rules) is averaged to create a unique value that indicates the level of account quality. Once defined, subsequent data improvement runs would re-calculate these values, creating an historical set of metric data that truly shows the improvement in data – now a measurable corporate asset.
Interactive Data Entry
As a result of automatic data improvement, there will inevitable be exceptions that require human intervention. A highly useable and effective data entry module enables operators to correct exceptions and complete the process of data improvement. The operator is presented with one screen where all exceptions are passed to. Once the data entry is complete, the data is ready for export.
Reporting
Integrity incorporates Crystal Report to display and print data quality reports. The reports analyse the information improved in all the data quality plans, jobs or rules executed. This gives complete visibility of how the data has been manipulated and assess the quality of data improvements undertaken.
Reports can be exported into PDF, MS Word or MS excel formats.
International Product
Integrity uses Unicode to hold character information, hence, European and international languages are supported in this product. All the rules are designed to with.
|
To discuss how we can help please contact us:
|
To request a demonstration of our flagship products
 |
Find our most popular resources more easily than ever.
|
    |
