Characteristics of the Italian language in the AX NLG platform

Fundamentals

In Italian, you need to know the gender of the noun in order to form determiners, adjectives and numerals correctly. There are two genders: masculine and feminine. And there are two numbers: singular and plural. As for other Romance languages there is only one case for nouns: nominative. But for personal pronouns Italian has additionally accusative, genitive and dative.

grammatical namevaluesexamples
genderfemininela casa nera (the black house)
masculineil porto nero (the black port)
numbersingularuna grande casa (a big house)
pluraldue grandi porti (two big harbors)
cases (noun)nominativecasa (house)
cases (pronoun)nominativeMario sta costruendo una casa. Lui sta costruendo una casa.
(Mario is building a house. He is building a house.)
accusativeMaria sta costruendo una casa. Maria la costruisce.
(Maria is building a house. Maria builds it.)
genitiveC'è del formaggio? Ne voglio una fetta. (Is there any cheese? I want a slice of it.)
dativeMaria regala un libro a Luigi. Maria gli regala un libro.
(Maria gives Luigi a book. Maria gives him a book.)
adjective positionsbefore noungrandi porti (large ports)
after noun (default)casa rossa (red house)
verb tensespresentlui parla (he speaks)
imperfectlui parlava (he talked)
gerundparlando (speaking)
futurelui parlerà (he will speak)

The standard order of a noun phrase in Italian is the following: preposition + determiner + numeral + noun + adjective.

See for example:

su     questi  tre    libri      popolari
about  these   three  books[pl]  popular
PREP   DET     NUM    NOUN       ADJ
"about these three popular books"

Lexicon

Nouns

For Italian nouns, the lexicon needs to encode gender, preposition changes and if a determiner is present. If the lexicon entry is missing, the NLG platform tries to find the most probable gender based on heuristics.

Examples

The lexicon entry for porto (port) contains

  • gender: masculine
  • inflection table for case and number:
SingularPlural
Nominativeportoporti

The lexicon entry for Australia contains:

  • gender: female
  • inflection table for case and number
  • remove the definite determiner for the preposition in

And the lexicon entry for Malta contains:

  • remove all definite determiners
  • inflection table for case and number
  • replace preposition in with a

note

If you need lexicon entries for countries please write the support about that and you will get them for Italian with automatic handling of determiners.

Adjectives

For adjectives the lexicon can encode gender for singular and plural and the adjective position (before the noun or after the noun).

For position, the default is after the noun. When an adjective should be before the noun, this has to be encoded in the lexicon.

The most important cases (e.g. grande, buono, etc.) are already annotated as before nouns, but others may need a lexicon entry.

Verbs

The most common verbs are encoded in our software. If a verb inflects the wrong way, you should add it to the lexicon.

Container settings

Determiner

The AX NLG platform supports the following determiners for Italian: indefinite, definite, demonstrative, distal, possessive, and quantifier (every)

Numerals

The noun will automatically agree with the numeral number when you use a numeral variable. There is no need to add additional branches for numeral. Four types of numerals are possible on the AX NLG platform: cardinal, cardinal as digit, ordinal, and ordinal as digit.

cardinalordinal
textNove topi nella mia stanza.
(Nine mice in my room.)
Il noni giorno di scuola.
(The ninth day at school.)
digit9 topi nella mia stanza.
(9 mice in my room.)
Il 9i giorno di scuola.
(The 9th day at school.)

For Italian, ordinal numerals are written out until 20, otherwise (above 20) the output is in digit form. Cardinal numerals are written out until 30, otherwise (above 30) the output is in digit form. Take cardinal numerals for example:

trenta auto
(thirty cars)
31 auto
(31 cars)

Prepositions and determiners: contractions

On the AX NLG platform prepositions are automatically joined with determiners for the following prepositions: a, da, di, in and su.

Determiners and preposition may also be changed separately. For example the preposition a changes to ad before vowels like in ad Andorra and determiners like la or una are contracted with an apostrophe like in l'amica.

In the below sentence, the preposition da is joined with the female definite article la and the article is adapted to the following vowel a to dall‘.

Stavano viaggiando dall'Australia a Parigi.
(They were traveling from Australia to Paris.)

On AX NLG platform, when the settings for the container dall‘Australia are: preposition="da" and determiner=definite, the lexical information about gender and the vowels following each other (dalla A -> dall'A) results automatically in correct inflection.

Determiner switch

On the AX NLG platform, the noun container is used in the same way for both examples (Australia and Emirati Arabi Uniti): preposition="in" and determiner=definite.

Vivono in Australia.
(They live in Australia.)
Vivono nelle Emirati Arabi Uniti.
(They live in the United Arab Emirates.)

The lexical information changes the results, in other words, it switches the determiners. In the first example, "without definite article" is automatically applied based on the lexicon. In the second example,the inserted noun is set to be always plural (plurale tantum), and therefore in + definite article is contracted to "nelle".

Preposition switch

Another case here is the preposition switch from in to a for (mostly) islands. Let's use Malta as an example here. The sentence would be:

Vivono a Malta.
(They live in Malta.)

so the preposition changes to a and there is no determiner. So Malta has no gender in the lexicon, but two switches: replace preposition in with a and they don't use any determiners with prepositions.

Defaults: Some best practices for Italian nouns with switches

  • If there are definite determiners (e.g., for country names) possible for a noun phrase they should be activated for the container and then the determiner is removed via the lexicon where appropriate. This assumes "with definite article" is the default.
  • The default for the preposition switch (only used for in -> a in Italian) is in as in most other languages.