TECH: AI Project Proposal

[Nick Nicholas <nsn@MULLIAN.EE.MU.OZ.AU>]

I am *not* asking any of you to do my homework for me! :) (In fact, I
ask that you not). BUT: I've decided to do a semester project on a
Lojban to Prolog translator, and would ask your comments on the following
proposal. And if possible, make your comments soon: I want to have this
proposal on my supervisor's desk first think Monday morning (Sunday
afternoon, US Time; Sunday evening, UK Time). I'd particularly welcome
suggestions on what parts of Lojban grammar I should admit: I suspect
the program presented here is too ambitious for the 60 hours of work I
had in mind.

[cut here]

Project Proposal for 433-603: A Lojban-to-Prolog translator.

Lojban is an artificial language intended for human use, of the type
exemplified by Esperanto and Interlingua. It is an offshoot of an
earlier such language, Loglan, and shares with it the ostensible _raison
d'etre_ of helping test the Sapir-Whorf hypothesis by deviating from
natural human language in a known, predictable manner, and observing
whether this deviation would have a noticeable effect on its speakers'
thinking. The hypothesis and the merits of its testing in this manner
are not held in universal esteem in the Lojban community; I, for one, am
quite skeptical that such a project can or need be realised. What is
of interest here is that, in order to implement this 'deviation', the
languages have been explicitly based on predicate logic. Predicates
serve the role of verbs, predicates with determines preposed serve the
role of nouns, and predications serve as sentences.

The immediate question that arises on considering such a language is:
to what extent can a logic (not necessarily First Order) adequately
express what human language can express? Again, this is a matter of
contention within the community; I believe that, if Lojban acquires a
true (second-language) speech community, its speakers will end up speaking
a human language no matter what, and that in the conflict between logic
and human language instinct, the latter will win (this is complicated
by the considerable risk that the language will end up as a code for
English, with little autonomy from that language; this tendency has been
resisted so far). Nonetheless, the language already provides a model
for natural language, with considerable expressive power, and with an
affinity to "logical forms" through its predicate logic origins, that
make a Lojban-to-Prolog translator an appealing task. The translator,
given a text in Lojban, would construct a Prolog database storing the
information denoted in the text.

The task is simplified in that an unambiguous context free grammar exists
for the language (implemented in YACC, with some imaginative use of error
recovery, but retaining LALR(1) nature). But even though syntactical
considerations have already been dealt with, most of the semantic issues
complicating logic-programming-based knowledge representation of natural
language remain in Lojban: higher-order predicates; metalinguistic comments
and attitudinals; the ambiguous semantic relationship of head and modifier
in word compounding; the representation of numbers, prepositional phrases,
non-logical connectives, negation, tense and modality; the distinction
between "the" and "a", partly echoed in the language's veridical and
non-veridical determiners; the distinction between individual and
collective plurals; subject-raising; relative clauses, and so forth.

In effect, a Lojban-to-Prolog translator would be addressing many of
the current issues in NLP knowledge representation, though it would be
biased towards predicate logic in the way it does so. With the way Lojban
grammar is structured, results will become available a short time into
the project without being distracted by parsing issues or syntactic
ambiguity.

In order to keep the project manageable, a subset of the language will have
to be considered; this is in line with the Lojban Canonicaliser proposed
by John Cowan. The subset of Lojban I intend to work on is be described
as follows, in stages:

1. Simple predications with a known predicate, and with arguments without
internal structure (Proper names, logical variables). No quantification
other than existential.
eg. mi prami da --- There exists an X such that LOVES(i,X).
2. Non-Veridical arguments (cf. English "the") based on predicates, with
internal arguments.
eg. mi catra le prami be le pulji --- KILLS(i,x) & LOVE(x,y) & POLICE(y):
I kill the lover of the policeman.
3. Veridical arguments (cf. English "an") based on predicates, with
internal arguments.
eg. mi catra lo prami be lo pulji --- There exist X and Y such that:
KILLS(i,X) & LOVE(X,Y) & POLICE(Y): I kill a lover of a policeman.
4. Resolution of logical connectives.
eg. mi nelci do .e ko'a ---> mi nelci do .ije mi nelci ko'a ---
LIKES(i,you) & LIKES(i,x1): I like you and him.
5. Restrictive and non-restrictive relative clauses.
eg. mi nelci le prenu poi do xebni ke'a --- (There exists x such that
HATES(you,x)) & LIKES(i,x) & person(x): I like the person you hate.
6. Higher order predicates.
eg. lenu mi cadzu cu nandu --- DIFFICULT(event:WALKS(i)): My walking is
difficult.
7. Prepositional phrases (other than tense and location).
eg. mi naumau do nelci ko'a ---> mi zmadu do leni da nelci ko'a ---
EXCEEDS(i,you,quantity:LIKES(X,x1)): I like him more than you.
8. Attitudinals.
eg. mi .ui sidju do ---> mi sidju do .ije mi gleki mi va'o lenu mi sidju
do: HELP(i,you) & HAPPY(i,i) & CONTEXT((state:HAPPY(i,i),event:HELP(i,you)):
I *smile* will help you, I am happy to help you.
9. Tense (including location), and prepositions of tense (including location).
Also includes modality and event contours.
eg. mi ba'o tavla ---> lenu mi tavla cu ba'o zei balvi zo'e:
AFTERMATH(event:talk(i,_,_,_),_): I have spoken.
10. Non-logical connectors.
eg. la gilbrt. joi la salivn. cu finti la mikadon. --- INVENT(X,mikado) &
JOINT_MASS(X,gilbert,sullivan): G & S (as a joint unit) wrote The Mikado.
11. Masses and sets as arguments.
eg. loi remna cu sipna: the mass of humans sleep (Even though it is not
true at any given moment that For all X: HUMAN(X) => SLEEPS(X)
12. Quantification (including numerical):
eg. mu le ze mensi cu cucycau: five of the seven sisters are barefoot.
13, Negation. Contradictory, scalar. Use of prenexes.
mi naku ro prenu cu prami: NOT(For all X:PERSON(X), LOVES(i,X))
mi ro prenu na prami: For all X:PERSON(X), NOT(LOVES(i,X))

Sections of Lojban Grammar not anticipated to be included in the model:

1. The mathematical subgrammar of Lojban.
2. Any analysis of word compounds.
3. Metalinguistic comments.

The detail of coverage of some sections, particularly tense, will probably
have to be curtailed due to time constraints. It is anticipated to have
this project take at most 80 hours of work.

Enclosures: [Bits of JL, John Cowan's sketch of a Lojban Canonicaliser]

Momenton senpretende paseman mi retenis kaj # [Victor Sadler, _Memkritiko_ 90]
   kultis kvazaux                           &  (NICK NICHOLAS. Melbourne.
      senhorlogxan elizeon                  #   Australia. IRC: nicxjo.
         (Dume:                             &   nsn@munagin.ee.mu.oz.au .)