Business Register in Sweden Public Register Statistical Register

Business Register in Sweden Public Register Statistical Register Study visit Ukrain 2012 -10 -15 1

Business Register Public part Administrative part Statistical part The administrative part consists of Legal Unit and Local Unit The public part is a subset of the administrative part The statistical part consists of Enterprise, Local Unit and other statistical units Study visit Ukrain 2012 -10 -15 2

FDB - Datamodel Study visit Ukrain 2012 -10 -15 3

Updating process and sources • • Three ways to update Picture of the updating process Sources and source priority Aggregation Study visit Ukrain 2012 -10 -15 4

Three ways to update: Batch (when updating many rows at the same time) Application (using the interface to update single rows, maintenance-work by the staff) Ad Hoc (special updates, corrections etc. Very restricted use) Batch updating will be described theoretically on the next slides Update online using the application will be shown practically at the end of the session Ad Hoc You update directly against the database without any support and control update a set Employees = 10 , Employee. Size. Class = 4 , Employee. S = ’TAX’ , Employee. SD = ’ 2012 -09 -15’ , Employee. N = 55 , Employee. ND = getdate() , Date. From = getdate() from Lo. U a where Lo. UId = 10124420 exec aggregation-module Study visit Ukrain 2012 -10 -15 5

Updating process batch Sources Tax Input tables Legal Unit Adress Company Questionnaire I_Le. UNace Next slide Presystem Local Unit Other I_Lo. UNace Other Study visit Ukrain 2012 -10 -15 6

Updating process batch (continue) Editing - address - phone - derive municipality code - assign source information Checking - against code tables - foreign key constraints Updating starts Source priority Ranking of Nace Database is updated using triggers and stored procedures Historical data When to aggregate Assign between units Insert Update Aggregation - legal unit - kind of activity unit - enterprise Study visit Ukrain 2012 -10 -15 7

Source variables All important variables should have source variables, for example Phone, char(11) - Source (Tel. K, char(3)) - Source Date (Tel. Ktid, smalldatetime) - Notifier (Tel. N, smallint) - Notification Date (Tel. Ntid, smalldatetime) - Verification Source (Tel. KB) - Verification Source Date (Tel. Ktid. B) - Verification Notifier (Tel. NB) - Verification Notification Date (Tel. Ntid. B) Study visit Ukrain 2012 -10 -15 8

Priority between sources datamodel A lower rank of a source will change a higher rank of a source if new source date is newer then protected date for the old value A higher (or equal) rank of a source will change a lower (or equal) rank of a source if source date is newer then old source date independent of protection date Study visit Ukrain 2012 -10 -15 9

Aggregation Updates in the local unit table could affect other units depending on what is updated. The following events affects • change of status of activity • change of number of employees • change of activity code In case of any changes as above the id of the aggregation level is inserted in an internal system table. All id in this table will be aggregated when aggregation starts Local unit is aggregated (assigned) to local kind of activity unit Local unit is aggregated to legal unit Local unit is aggregated to enterprise Local kind of activity unit is aggregated to kind of activity unit Study visit Ukrain 2012 -10 -15 10

THANK YOU VERY MUCH Study visit Ukrain 2012 -10 -15 11