Named entities normalization - an example

Cvetana Krstev, Jelena Jaćimović, Duško Vitas

In a sample text following named entities were recognized, normalized and XML-tagged:

• monetary expressions, tag <MONEY>, with attributes:
  • VALUE with normalized value — normalized numerical value and the ISO normalized currency unit;
  • MOD with possible modifications of a recognized value:
    • EXACT
    • LESS_THEN
    • GREATER_THEN
    • APPROX
• measurement expressions, tag <MEASURE>, with attributes:
  • VALUE with normalized value — normalized numerical value and a normalized measurement unit (according to SI and USA recommendations);
  • MOD with possible modifications of a recognized value:
    • EXACT
    • LESS_THEN
    • GREATER_THEN
    • APPROX
• temporal expressions, tag <TIMEX3>, with attributes:
  • TYPE type of a temporal expression:
    • DATE
    • TIME
    • DURATION
  • VALUE with normalized value according to ISO 8601-1997 standard;
  • MOD with possible modifications of a recognized value:
    • MID
    • ON
    • START
    • END
    • APPROX