Skip to content

Execute Output

At the end of the Execute phase, there is one output. It contains the best matchings in ascending order (with the codes only), their rank, the level metric, the concat metric and a measure, “METRIC”, that represents the overall goodness of the match, calculated as the average between the median levels distance and concat distance.

Appendix 1: Input and Output Data Sets

Data Set for the Train Phase:

  • one input data set
  • Data Sets for the Execute Phase:
  • one input data set
  • one output data sets

An array can define some fields, meaning more data set columns can be associated with the field.

Appendix 1.1: Input Train Data Set

This data set contains the codes and descriptions for the levels (punctually compared, without the concatenation of the strings) and the concat (concatenated before the comparison). The code fields are arrays. For each description field, there must be one code field only.

  • LEVEL_CODE_0: code of the first level description field
  • LEVEL_CODE_1: code of the second level description field
  • …
  • LEVEL_DESCRIPTION_0: first level description field
  • LEVEL_DESCRIPTION_1: second level description field
  • …
  • CONCAT_CODE_0: code of the first description field to be concatenated
  • CONCAT_CODE_1: code of the second description field to be concatenated
  • …
  • CONCAT_DESCRIPTION_0: first description field to be concatenated
  • CONCAT_DESCRIPTION _1: second description field to be concatenated

For example:

LEVEL_CODE_0 LEVEL_DESC_0 CONCAT_CODE_0 CONCAT_DESC_0 CONCAT_CODE_1 CONCAT_DESC_1
E_1 USA A11 HOME EXPENDITURE CC_01 UTILITY
E_1 USA A11 HOME EXPENDITURE CC_05 HEAT
E_3 APAC A11 HOME EXPENDITURE CC_09 GAS
E_1 USA B12 FURNITURE CC_03 TABLES

Appendix 1.2: Input Execute Data Set

The structure of the Execute Input Data set is the same as the Train one. The number of level codes and description must be the same; the number of concat codes and description can differ.

For example:

LEVEL_CODE_0 LEVEL_DESC_0 CONCAT_CODE_0 CONCAT_DESC_0
NA NORTH AMERICA CD1 SEWER
NA NORTH AMERICA CD2 HEAT
AS ASIA CD2 HEAT
NA NORTH AMERICA CD3 GAS

Appendix 1.3: Output Execute Data Set for Pred

This data set contains the matching code fields between train and execute, the matching rank, and the distances measure.

  • LEVEL_CODE_0: code of the first Execute level description field
  • LEVEL_CODE_1: code of the second Execute level description field
  • ..
  • LEVEL_DESCRIPTION_0: first level description field
  • LEVEL_DESCRIPTION _1: second level description field
  • …
  • PRED_LEVEL_CODE_0: code of the first Train level code field
  • PRED_LEVEL_CODE_1: code of the second Train level code field
  • …
  • CONCAT_CODE_0: code of the first Execute concat description field
  • CONCAT_CODE_1: code of the second Execute concat description field
  • …
  • CONCAT_DESCRIPTION_0: first description field to be concatenated
  • CONCAT_DESCRIPTION_1: second description field to be concatenated
  • …
  • PRED_CONCAT_CODE_0: code of the first Train concat description field.
  • PRED_CONCAT_CODE_1: code of the second train concat description field
  • …

For example:

  • RANK: rank order of the matching
  • LEVELS_METRIC: the median of the vector distances of the level fields (for each match)
  • CONCAT_METRIC: the vector distances of the concat fields
  • METRIC: the Average between the LEVEL METRIC and the CONCAT METRIC

Below you can see a schema of how the AI Automapping model works*:

![Immagine che contiene testo, schermata, Carattere, software

Descrizione generata automaticamente](../../../assets/img/72e2b06b3a_AutoMap_Execute-Output.png)

Appendix 2: Parameters

The parameters to set for the Train phase include the following:

  • LOG LEVEl:

  • DEBUG (all information logged)

  • INFO (all warnings, errors, critical errors, and general information logged without debug information)
  • WARNING (all warnings, errors, and critical errors logged)
  • ERROR (all errors and critical errors logged)

  • CRITICAL (only critical errors logged)

    Logging messages that are less severe than the level specified for the parameters are ignored; logging messages with a severity level or higher are emitted. - TRANSLATE BEFORE: Whether to translate texts before using the NLP models (Y) or directly leverage multilingual models (N)

For example:

The parameters to set for the Execute phase include:

  • LOG LEVEL:

  • DEBUG (all information logged)

  • INFO (all warnings, errors, critical errors, and general information logged without debug information)
  • WARNING (all warnings, errors, and critical errors logged)
  • ERROR (all errors and critical errors logged)

  • CRITICAL (only critical errors logged)

    Logging messages that are less severe than the level specified for the parameters are ignored; logging messages with a severity level or higher are emitted. - NUMBER OF CLASSES: number of matchings returned for every Execute input row. The default value is 1.

  • TRANSLATE BEFORE: Whether to translate texts before using the NLP models (Y) or directly leverage multilingual models understanding (N)

  • PRIORITIZE LEVELS: Whether to provide the final ranking by prioritizing the level matches as they are ordered (Y) or not (N)

For example: