US 20090083619

  1. A method in a data processing system, comprising the steps of: receiving a document containing numerical values; receiving indications of characteristics of the numerical values, the characteristics including a unit and a magnitude; and adding the received indications into the document as tags associated with the numerical values to create a markup document.

This is very similar to "named entity recognition": http://en.wikipedia.org/wiki/Named-entity_recognition

Named entity recognition searches a text for known strings or string patterns, e.g. gene names (like "bco2"), and attaches markup (in other words: tags) to them in the text. You could use the named entity recognition functionality of the open-source tool KNIME to construct a recognizer that recognizes string patterns that represent numbers (using regular expressions) and search for the "characteristics" (e.g. unit) as classic named entities.


  • this claim is about attaching semantics to numerical values found in a text.
  • there's lots of ways this can be done, for example with microformats: example (from: http://microformats.org/wiki/measure):

" The action figure has a height of 5 in."

=> markup:

<p class="hmeasure"> The <span class="item">action figure</span> has a <span class="type">height</span> of <span class="num">5</span> <span class="unit">in</span>. </p>

  • Does your example show the classes being inserted automatically? – George White May 16 '14 at 23:09

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.