Illustrates Data Ireland ability to project manage and process large complex projects.
In Q1 2007 Data Ireland was contracted to carry out the data processing of Electoral Register data for the General Election campaign, specifically the distribution of polling cards. This project included:
Data Receipt – coordination of supply of over 3.2 million records from 42 different sources (each Local Authority). This involved managing each Local Authority to ensure data was received in time to meet tight campaign deadlines, validation of data on receipt and organising a re-supply of records where the data did not conform to agreed format and/or content.
Data Pre-processing – The complex nature and limitations of the data source meant that each individual name and address record was supplied across 14 different files with each file containing a different component of the name, address, polling area or other associated data. This process was further complicated by data inconsistencies within and between data sources. The initial pre-processing stage involved compiling a composite database containing all data elements in a uniform format.
Cleansing and Validation – Each name, address and polling area data supplied was then validated against reference tables. This process has several aspects:
- Validation of name data supplied against our comprehensive reference tables
Validation of address data supplied against our comprehensive reference tables (includes aliases and common miss-spellings) and against the GeoDirectory
- Automated enhancement of data where errors/omissions have been identified
- Automated verification of addresses against a reference table of Polling Stations.
- Manual review and enhancement of data where errors/omissions have been identified
Suppressions and Amendments - Client-specific suppressions and amendments were briefed on an ad-hoc basis. These included changes to Polling Station allocations of addresses as well as specific amendments to individual name and address records.
Sortation – to ensure efficient distribution of the mailing the data was sorted initially according to standard Postaim requirements. Because of the large volume a further sortation was run to sort addresses in alpha/numeric order within each Postaim code.
QC checks – Thorough QC checks were carried out on all data prior to output. These QC checks ensure that data quality and the application of Polling Station and postal sortation codes are accurate. A combination of automated cross-referencing checks and manual review of data is employed at this stage.
PGP encryption was employed to transfer all mail files. Upon receipt, produced a series of proofs for verification against the original data supplied prior to final printing and finishing.
Status Meetings – During the course of the project a series of Status Meetings were organised to ensure the project was completed accurately and to deadline.