Part-Of-Speech (POS) tagging

Part-of-speech tagging (POS) is the process of assigning a POS (noun, adjective, verb, etc) to a word. In Wowool you can choose, whether you want to preserve all the POS when there is not enough evidence to assign a single POS to a word, or (the default) to run it as a deterministic process where every word will get the most likely POS.

The POS are used to simplify rule writing, as they provide another layer of abstraction: Nouns (Nn) are usually used to express concepts, like “mother” or “freedom”, Verbs (V) denote action, like “collect” or “jump”, adjectives (Adj) qualify nouns: “rich”, “mysterious”, “yellow”.

Literal    Lemma        POS
"Age"      'age'        Nn-Sg
"is"       'be'         V-Pres-Sg-be
"an"       'a'          Det-Indef
"issue"    'issue'      Nn-Sg
"of"       'of'         Prep-of
"mind"     'mind'       Nn-Sg
"over"     'over'       Prep-Std
"matter"   'matter'     Nn-Sg
"."        '.'          Punct-Sent

The POS tags can either be used fully (eg “V-Pres-Sg-be”), or as so-called umbrella tags (eg “V”).

Umbrella tags are simplified POS tags containing the POS until the first hyphen. Umbrella tags are consistent for all languages, whereas the full tags differ.

Umbrella Tag	Meaning	Example
Adj	Adjective	“big”, “incendiary”
Adv	Adverb	“incredibly”
Aux	Auxiliary	“do”, “will”
Conj	Conjunction	“if”, “and”
Det	Determiner	“the”, “my”
Interj	Interjection	“yes”
Nn	Noun	“car”, “imagination”
Num	Number	“1”, “forty”
Part	Particle	“to”, “not”
Prep	Preposition	“in”, “for”
Pron	Pronoun	“he”, “anybody”
Prop	Proper Noun	“England”, “Michelle”
Punct	Punctuation	“,”, “:”
V	Verb	“is”, “came”

What’s the use of Part Of Speech

POS is used to generalize: it helps us to write rules in a simple way: for example, if we are looking for descriptions of wine, the adjectives can help us:

rule:
{
    Adj 'wine'
} = WineType;

Output:

red wine

white wines

delicious wine

tasteless wine

It also helps us to disambiguate or distinguish a word’s meaning: for example ‘store’ can be a place where merchandise is offered for sale (a noun) or an action (a verb): ‘to put something in a place where it is available’. The same happens with the word ‘terminal’: it can be an adjective in ‘a terminal illness’ or a noun ‘passengers can proceed to the terminal’.

If you want to annotate places, you could use the POS to find the correct meaning:

rule:
{
    (
        <'store',Nn>
    |   <'terminal',Nn>
    )
} = Place;

Note

POS can only be referred to in rules, not in lexicons, as you will see in the following chapters.