Natural Language Processing

Morphology. As an empirical matter, take the morpheme "ness." It is appended to e.g. "dark". This linguistic structure is one in which the prepended morpheme must have a complement. Dark-Light. Darkness not Lightness. Now, we take the word that came before. Is it a monomorphemic word? If it is, such as a determiner, then what information does it give us? Moving beyond morphemes, we look at word-level information retrieval. A string of parts of speech. We number each part in context so that we determine the primary word in the sentence, then the words wich must, by nature, make it relevant. etc etc etc...