AtlasDedupe

AtlasDedupe

Here you can find the most frequently asked questions forĀ AtlasDedupe

AtlasDedupe is a highly configurable product that accommodates the needs and requirements of many of its users through a single graphical user interface. The AtlasDedupe configuration employed will depend largely on what you are looking to achieve by your run. It may be that you are proposing a mailing whereby you want one mailpiece delivered per household. In this instance, you would use only AddressKey and ExtendedKey. If you want to mail to multiple prospects or customers potentially within the same household, you would need to include the NameKey into the equation.

AtlasDedupe is best utilised with Hopewiser AddressKey/NameKey/ExtendedKey, etc. It is possible to select 2 user defined keys, but you are unlikely to get the same comprehensive coverage offered by the Hopewiser keys using just these, and these user defined keys are unlikely to accommodate things like common misspellings, etc, which the Hopewiser keys, in the main, do.

The AtlasDedupe passes facilitate differing levels of file deduplication. The different criteria (which parts of which keys the pass uses to establish a duplicate match) are displayed in the Pass Details frame towards the bottom of the Select Passes window. Which pass/es you choose to utilise, again, depends on what you are looking to achieve, and the deduplication keys you have managed to create in the pre-deduplication address matching run.

The AtlasDedupe Report panel allows you to define what appears in the landscape report output which illustrates the duplicate blocks match for each successive pass. It is recommended that you include the fewest amount of fields from your input that are pertinent to you (input name and address maybe), otherwise the output will overrun the end of the page and the most pertinent information may be lost. This report can be used to assess the validity of duplicate matches found and present the user with ideas on identifying trends in duplication.

The AtlasDedupe Filter panel allows the user to filter the eventual output file obtained after running the process. It may be that you want only unique records within a file or files. If this is the case, then you would just have the Unique check box within the Filter screen selected. It may be that you only want duplicate records from input file C (file 3 of a multi-file input) output. This is possible using the Filter panel also.

The AtlasDedupe Merge panel allows the user to extract data from duplicate records and 'promote' these up into the first record of that duplicate block (the one that will be retained) before dropping the duplicate records. This means that you can drop duplicate data (invariably based on address and name information) without any loss of associated data (invoice numbers, order numbers, for example).

This is a non-standard piece of functionality which allows a user to step through duplicates detected automatically and gives the user the opportunity to alter deduplication sequences, flag as non-duplicates, etc, basically confirming the automatic matches made by AtlasDedupe.

This facility allows the user to specify the match accuracy required, within passes not automatically set to be 100%, so that more duplicate matches can be achieved. This would usually allow, one or more characters difference between successive keys. The lower the Match Accuracy setting, the more character differences allowed in flagging records as duplicates of others. This can be useful in, say, a deduplication which includes an input name where Robert and Bob are the same person with, in reality, a different forename initial cited in the two different records.

AtlasDedupe uses 3 temporary paths in performing a deduplication on a file or files. Where possible, if you can set these to 3 separate disks (local disks; it is not prudent to use remote network drives), this will increase the performance of AtlasDedupe in terms of processing speed.

This tab fine tunes a run which includes a Pass 0. It is here where you can specify whether you want to include AddressKey, NameKey, CompanyKey and/or the user defined keys (2) within this 'pass'. A user defined key can be any simple string. You can also eliminate certain elements of the ExtendedKey if they are irrelevant to your matching of duplicates by unchecking that item.

Within AtlasDedupe, this implies that a key required for a deduplication 'pass' selected (required fields are cited on the right-hand side of the Input Mappings window) has not been linked. Recheck the AtlasDedupe Input Mappings settings and ensure that all required fields are 'linked' to the corresponding input file field.

Check your Input Mapping panel settings first of all. The likelihood is that one or more of the 'required' fields is mapped incorrectly. False duplication matches can also be a result of setting the Match Accuracy figure on the main AtlasDedupe screen too low. This will mean that passes not automatically set at 100% will be 'allowing' many more character differences than prudent.

For further details on AtlasDedupe and how to best utilise it, please refer to the PDF documentation within the application installation directory. If you have any questions not covered by this document, please contact Hopewiser technical support department on 0161 924 2801.