Can I use QAS Pro as a Batch Address Cleaning Engine?

Document type: Tech doc  
Associated product(s): Product name: BatchImplementation: API (more products)
Product name: ProImplementation: API

Views: 605  |  Created: 2 years agoLast updated: 2 years ago

Can I use QAS Pro as a Batch Address Cleaning Engine?

Introduction

A common question often asked by Experian QAS customers is whether QAS Pro API can be used as a batch-cleaning engine? In theory it would be possible to call the QAS Pro API repeatedly, each time passing in a complete address record from a database or file, however the results produced from such use would fall far below the level that most people would expect and accept.

This in no way reflects any lack of functionality with the QAS Pro API application, and instead highlights the need to have a unique solution to verify and cleanse existing addresses. This document thus highlights the intended functionality of QAS address capture and cleansing products.

Why Does QAS Pro API not work in a batch-cleansing mode?

QAS Pro API is designed to capture a complete address, by performing a search on a minimum amount of information supplied by the user. If that information is not sufficient to identify a single delivery point (i.e. a letterbox), then a list of the choices (e.g. a range of premise numbers) is generated for the user to choose from, to ‘drill down’ to the correct address via a pick list. An example of how such functionality performs in batch mode follows.

If an address record is taken from a database and passed to the QAS Pro API search engine, a single search on all of the information will be initialised. Should this information not be sufficient to return a single address, a pick list will be generated. This could be handled in several ways:

1. The pick list could be displayed to a user, however this would defeat the point of a batch process, and would also greatly add to the time and other resources required.

2. Alternatively, it would be possible to automatically note that the record is incorrect and proceed to the next, thus marking the record for further work, but the problem would not actually have been solved.

In reality quite a high percentage of stored address data is not complete, or contains contradictory elements (e.g. incorrect postcode), and as the above example illustrates, without user intervention the QAS Pro API application is not suitable for correcting such mistakes in an efficient manner.

Quite simply QAS Pro is intended as an interactive address-capturing product, whose many benefits are best exploited when driven by a user. An alternative solution to such problems is now discussed.

So what is the alternative solution?

In the same way that QAS Pro was designed for interactive use, there also exists an API to work in automatic mode, to verify and cleanse existing address records. This application is called

QAS Batch API

QAS Batch API is designed specifically for taking an existing address record, pre-processing it, and then potentially performing multiple searches on different combinations of address elements, each being intelligently selected. The results of these searches are then compared automatically, so as to ascertain the exact errors within the existing address.

Further to that, by use of configuration settings pre-determined by the user, QAS Batch API can often correct those errors within the rules that have been set, as well as providing comprehensive reporting on each record, and overall statistical information in a report format.

Below is an explanation of how data gets processed within the QAS Batch API process.

1. Pre Processing

Existing address information is processed to identify basic elements, and convert abbreviations etc.

2. Build Search List

The address elements are examined, and appropriate searches constructed.

3. Verification Search Engine

This operates at a lower level than the QAS Pro API search engines, and performs the required searches.

4. Comparison

The results from the searches are compared, so as to ascertain the exact errors, and ensure that the suggested PAF address is appropriate.

Summary

By creating two separate products to perform these two very different tasks, Experian QAS have been able to tailor the functionality of each product so as to achieve optimum performance using individually integrated address capture and address cleansing solutions.

Was this document helpful?

What can we do to improve this information?