Anchoring and extracting text that matches parts of speech

Currently FACTS does not model these in its framework as it does entities. However, you can use existing fact types and use the special part-of-speech tags that the advanced language packs (not the standard ones) use, as alt labels.

For example, if you wish to extract all numbers from some document, you could build an context using “Unsequenced In Document” and have a single concept fact, with the alternative label {CARD}, which stands for the cardinal number part of speech tag.

A full list of the part-of-speech tags that are available can be found both:

• in the framework under FRAMEWORK SETTINGS > Parts Of Speech
• and in the portal here Part-of-Speech tags