Data Quality user documentation
home/Address validation/Clean Web Service/Overview/Workflows

Workflows

The following workflows are available to be used in Clean Web Service.

Required settings

The **AddressData **and Layout settings must be used.

Processing overview

For each Record, the following processing will take place:

  • Match – the address is compared against the postal database for the AddressDataSet which is specified in the settings section.
  • Clean – If a match is found, any missing elements of the address are added, and any incorrect or out of date elements are fixed.
  • Enrich – Any DataPlus fields which are included in the Layout you specified in the settings section are appended to your Records.
Response overview

The first two strings in each of the response's Record tags contain the match profile (if you specified in the settings section that match profiles should be returned) and the match code for the record. These correspond to the quality of the match and the changes made to the record.

The next set of strings contain each line of the address, according to the Layout you specified in the settings section. Any blank strings indicate that a field is not used. For example, many addresses do not require the state/county field.

The final set of strings contain any DataPlus fields which are included in the Layout that you requested in the settings section.

An additional string will contain a reference field, if you requested one to be included, either by specifying it in the InputFieldMappings, or by setting ReferenceField to True.

Required settings

The InputFieldMappings setting must be used.

Processing overview

For each Record, the following process will take place:

  • Validate – the email address is validated in real time against the relevant service provider.
Response overview

The first string in each of the response's Record tags contain the email match profile (if you specified in the settings section that match profiles should be returned). This corresponds to the quality of the match and the changes made to the email address.

One of the following strings contains the cleaned and validated email address, if available. If the email address could not be validated, this string will be empty.

If you included a reference field, this will also be included as a string.

The order which the last two strings are returned in depends on the value you supply for the InputFieldMappings setting.

Required settings

The InputFieldMappings, AddressDataSet and Layout settings must be used.

Processing overview

For each Record, the following processing will take place:

  • Match – each postal address is compared against the postal database for the AddressDataSet which is specified in the settings section.
  • Clean – If a matching postal address is found, any missing elements of the address are added, and any incorrect or out of date elements are fixed. At the same time, each email address is checked for common errors, such as mistyped elements of the email address, and corrected.
  • Enrich – Any DataPlus fields which are included in the postal address Layout you specified in the settings section are appended to your Records.
  • Validate – Email address is validated in real time against the relevant service provider.
Response overview

The first two strings in each of the response's Record tags contain the match profile (if you specified in the settings section that match profiles should be returned) and the match code for the record. These correspond to the quality of the match and the changes made to the record.

One of the next set of strings contain the cleaned postal address, according to the Layout you specified in the settings section. Any blank address elements indicate that a field is not used. For example, many addresses do not require the state/county field.

Another set of strings contain any DataPlus fields which are included in the Layout that you requested in the settings section.

Another string contains the cleaned and validated email address, if available. If the email address could not be validated, this string will be empty.

If you included a reference field, this will also be included as a string.

With the exception of the match profiles and match code (which are always returned first) the order which the strings are returned in depends on the value you supply for the InputFieldMappings setting.

Processing and cleaning

When using a workflow which supports postal address cleaning, each successful request to Clean Web Service will result in a response which includes an AdditionalInformation object, containing fields similar to those shown below.

<additionalinformation>
  <namedvalue>
    <key>DataSetVintage</key>
    <value>21/02/2014</value>
  </namedvalue>
  <namedvalue>
    <key>BatchVersion</key>
    <value>7.30</value>
  </namedvalue>
  <namedvalue>
    <key>DatasetName</key>
    <value>GBR</value>
  </namedvalue>
  <namedvalue>
    <key>TotalRecords</key>
    <value>62</value>
  </namedvalue>
  <namedvalue>
    <key>RecordsProcessed</key>
    <value>60</value>
  </namedvalue>
  <namedvalue>
    <key>RecordsSubmitted</key>
    <value>62</value>
  </namedvalue>
  <namedvalue>
    <key>CLEAN-GBR_RecordsCleaned</key>
    <value>56</value>
  </namedvalue>
  <namedvalue>
    <key>CLEAN-GBR-GBRGRD_Dataplus_RecordsCleaned</key>
    <value>52</value>
  </namedvalue>
  <namedvalue>
    <key>CLEAN-GBR_RecordsProcessed</key>
    <value>60</value>
  </namedvalue>
  <namedvalue>
    <key>CLEAN-GBR-GBRGRD_Dataplus_RecordsProcessed</key>
    <value>52</value>
  </namedvalue>
</additionalinformation>

It is important to understand the meaning of these values, as they directly affect the amount you are charged for using Clean Web Service. Whether you are charged for the number of records processed or cleaned depends on your individual license agreement.

RecordsProcessed

When cleaning data, this is the number of records which were successfully processed (those that achieve a Match Success indicator between K and R).

When enriching data, this is the number of records for which DataPlus values were returned from a particular DataPlus set. Charges are made for each DataPlus set used, no matter how many individual DataPlus fields are returned. DataPlus sets are identified by their six-letter code. In the example above, the GBR Grid DataPlus set (GBRGRD) has been used.

In the example above, we can see that:

  • 62 records were sent to Clean Web Service (as shown by the value of RecordsSubmitted).
  • Of these, 60 records were processed against the GBR AddressDataset, which tells us that two records could not be matched successfully.
  • Of these, 52 records were processed against the GBRGRD DataPlus set. See RecordsCleaned below for an explanation of this.

The reason(s) for any record remaining unmatched can be determined by its individual match code, which will be the only non-blank string for any unmatched records.

RecordsCleaned

When cleaning data, this is the number of records which were cleaned with high or intermediate confidence (those that achieve a Confidence Level of 5 or 9).

When enriching data, this is the number of records which were cleaned with high confidence (those that achieve a Confidence Level of 9 only), for which at least one DataPlusField from the indicated DataPlus set was appended.

In the example above, we can see that:

  • 56 records were cleaned against the GBR AddressDataset, which tells us that every record which was matched successfully was also matched with intermediate or high confidence.
  • Of these, 52 records were cleaned against the GBRGRD DataPlus set. One possible explanation of this is that four of the records were matched with only intermediate confidence; it is also possible that no grid information could be found for these records. Analysis of the individual match codes will help to determine which is the most likely.
Clean Web Service

Overview