Words have meaningful sub-parts that aren't all words:

unfamiliarity un familiar ity

formalize form al ize

undeniability un deny able ity

Right	Wrong
deny deny able un deny able un deny able ity undeniability	deny un deny un deny able un deny able ity undeniability

walked
walk	ed
Stem	Suffix

walked

walking

walks

formalized
formal		ed
Stem		Suffix
form	al	ed

Two kinds of programs:

1. Morphological Analyzer: Not only takes a word apart into its component "atoms" (morphemes), but also produces the right "structure".
   Right: formalize = [[form + al] + ize]
   Wrong: formalize = [form + [al + ize]]
   
2. Stemmer: Finds stems
   formalize => formal
   formalized => formalize
   formalized => formal

Form Stem Present Participle walking walk Past Participle walked walk Past Form walked walk Present Tense walks walk

Form Stem breaking   break broke   break broken   break breaks   break

Form Stem bleeding   bleed bled   bleed bled   bleed bleeds   bleed

Inflectional morphology is about the different forms of a single word:

Inflectional Verb

break	->	breaks	3rd person present tense singular
		breaking	present participle
		broken	present participle
		break	3rd person present tense plural
		broke	past tense

Inflectional Noun

cat	->	cat	Singular
		cats	Singular

Inflectional Adjective

happy	->	happy	Positive
		happier	Comparative
		happiest	Superlative

Dervivational morphology is about making new words.

relational	->	relate
conditional	->	condition
digitizer	->	digitize
predication	->	predicate
hopefulness	->	hopeful
feudalism	->	feudal
formalize	->	formal
feudalism	->	feudal
allowance	->	allow
effective	->	effect

1. Let vowels be:
   A E I O U and Y (after non- A,E,I,O,U)
   try: "y" is a vowel
   bay: "y" is not a vowel
   
2. Let consonants be anything other than a vowel.
3. Let V stand for any sequence of vowels.
4. Let C stand for any sequence of consonants.
5. We can represent any word by the pattern:
   (C)(VC)^m(V)
   Parentheses here mean optionality. That is words can skip the initial consonant sequence C and start witha vowel sequence("at"), they can skip the VC sequence and go directly to a final vowel ("tree"), or they skip the final vowel ("cat").

   Here "m" is Kleene-operator m. We allow any number of VC sequences ( each a "syllable").

6. WE call the value of "m" for a particular word its measure.

m	the measure of the stem
*S	the stem ends with S (and similarly for other letters)
*v*	The stem contains a vowel
*d	the stem ends with a double consonant (e.g. -TT or -SS)
*o	the stem ends in CV1C, where the second C is not W, X, or Y (e.g. -WIL, -HOP)

1. Plural Nouns and Third Person Singular verbs (-s ending)

   cats	=>	cat
   walks	=>	walk

2. Verbal Past tense and progressive forms

   walked	=>	walk
   walking	=>	walk

3. Y => I
   happy => happi
4. Derivational Morphology I: Multiple suffixes (-TIONAL, -ENCY, -ALLY, -IZATION, -ALISM.etc.)
5. Derivational Morphology II: More Multiple suffixes (-ICATE, -ATIVE, .etc.)
6. Derivational Morphology III: Single suffixes (-AL,-ANCE,-ENCE, -IC,-ABLE)
7. Cleanup

No conditions for these rules: Only one rule per set may apply to a given word. The rule with the LONGEST MATCH should apply. This is what explains rule (c) below. Any word that matches (c) will not undergo (d).

	Rule	Example
(a)	SSES->SS	caresses => carress
(b)	IES->I	ponies=> poni ties=> ti
(c)	SS->SS	ponies=> poni ties=> ti
(d)	S->eps	cats=> cat

	Condition	Rule	Example
(a)	(m > 1)	EED->EE	feed => feed agreed => agree
(b)	(*v*)	ED->eps	plastered=> plaster bled => bled
(c)	(*v*)	ING->eps	try=> try spring => spring

	Condition	Rule	Example
(a)		AT => ATE	conflat(ed) => conflate
(b)		BL->BLE	troubl(ing) => trouble
(c)		IZ->IZE	siz(ed) => size
(d)	(*d & !(*L or *S or *Z))	CC -> C	hopp(ing) => hop tann(ing) => tan fall(ing) => fall hiss(ing) => hiss fizz(ing) => fizz
(e)	(m=1 & *o)	eps -> e	fail(ing) => fail fil(ing) => file

(*v*)

Y -> I

happy => happi sky => sky

Condition	Rule	Examples
(m > 0)	ATIONAL => ATE	relational => relate
(m > 0)	ENCI => ENCE	valenci => valence
(m > 0)	IZER => IZE	digitizer => digitize
(m > 0)	ABLI => ABLE	comfortably => comfortable
(m > 0)	ALLI => AL	radicalli => radical
(m > 0)	ENTLI => ENT	differentli => different
(m > 0)	IZATION => IZE	vietnamization => vietnamize
(m > 0)	ELI => E	vileli => vile
(m > 0)	OUSLI => OUS	analogousli => analogous
(m > 0)	ATION => ATE	predication => predicate
(m > 0)	ATOR => ATE	operator => operate
(m > 0)	ALISM => AL	feudalism => feudal
(m > 0)	IVENESS => IVE	decisiveness => decisive
(m > 0)	FULNESS => FUL	hopefulness => hopeful
(m > 0)	OUSNESS => OUS	callousness => callous
(m > 0)	ALITI => AL	formaliti => formal
(m > 0)	IVITI => IVE	sensitiviti => sensitive
(m > 0)	BILITI => BLE	sensibility => sensible

Condition	Rule	Examples
(m > 0)	ICATE => IC	triplicate => triplic
(m > 0)	ATIVE => eps	formative => form
(m > 0)	ALIZE => AL	formalize => form
(m > 0)	ICITI => IC	electriciti => electric
(m > 0)	FUL => eps	hopeful => hope
(m > 0)	NESS => eps	goodness => good