TCUK15 Speaker – Mike Unwalla to present on “Patterns in language for part-of-speech disambiguation”

Mike Unwalla will be presenting on “Patterns in language for part-of-speech disambiguation”.

Mike Unwalla

In the ASD-STE100 specification, some words are approved for only one part of speech. For example, the word ‘oil’ is approved as a noun, but it is not approved as a verb.

An effective term checker gives a warning only if ‘oil’ is a verb:
Correct, no warning: The oil was dirty.
Incorrect, warning: When you oil the bearing, you must…

How can software ‘know’ whether ‘oil’ is a noun or a verb?

Short answer: In the structure ARTICLE+X+BE, X is a noun.

Mike shows how software can use sets of simple patterns to identify the part of speech that a word has.