andre freitas
TRANSCRIPT
Symbolic AI Andre Freitas
Photo by Vasilyev Alexandr
Acknowledgements
• Based on the slides of:
– NaturalLI: Natural Logic Inference for Common Sense Reasoning
– Modeling Semantic Containment and Exclusion in Natural Language Inference. Bill MacCartney 2008: https://slideplayer.com/slide/5095504/
– NatutalLI. G. Agneli 2014: https://cs.stanford.edu/~angeli/talks/2014-emnlp-naturalli.pdf
This Lecture
• Natural Language Inference.
Text Entailment
• Does premise P justify an inference to hypothesis H?
• P : Every firm polled saw costs grow more than expected, even after adjusting inflation.
• H : Every big company in the poll reported cost increases.
• YES
– What if we change the quantifiers to Some?
Text Entailment
• Does premise P justify an inference to hypothesis H?
• P : The cat ate a mouse.
• H : No carnivores eat animals.
• NO
NLI: a spectrum of approaches
lexical/ semantic overlap
Jijkoun & de Rijke 2005
patterned relation
extraction
Romano et al. 2006
semantic graph
matching
MacCartney et al. 2006
Hickl et al. 2006
FOL & theorem proving
Bos & Markert 2006
robust,
but shallow
deep,
but brittle
natural logic
(this work)
Problem:
imprecise easily confounded by
negation, quantifiers, conditionals,
factive & implicative verbs, etc.
Problem:
hard to translate NL to FOL idioms, anaphora, ellipsis, intensionality, tense, aspect, vagueness, modals, indexicals, reciprocals, propositional attitudes, scope
ambiguities, anaphoric adjectives, non-intersective adjectives, temporal & causal relations, unselective quantifiers,
adverbs of quantification, donkey sentences, generic determiners,
comparatives, phrasal verbs, …
Solution?
more than expected, even after adjusting for inflation. 0.9 0.6 0.9 0.4 0.9 0.8
Shallow approaches to NLI
• Example: the bag-of-words approach [Glickman et al. 2005]
– Measures approximate lexical similarity of H to (part of) P
P Several airlines polled saw costs grow
H Some of the companies in the poll reported cost increases .
0.9
No
None
• Robust, and surprisingly effective for many NLI
problems.
• But imprecise, and hence easily confounded
• Ignores predicate-argument structure — this can be
remedied
• Struggles with antonymy, negation, verb-frame alternation
Relies on full semantic interpretation of P & H
(greater-than (magnitude g)
The formal approach to NLI
P Several airlines polled saw costs grow more than expected,
even after adjusting for inflation.
(exists p (and (poll-event p)
(several x (and (airline x) (obj p x)
(exists c (and (cost c) (has x c)
(exists g (and (grow-event g) (subj g c)
..... ?
• Need background axioms to complete proofs — but from
where?
• Besides, NLI task based on informal definition of inferability.
• Bos & Markert 06 found FOL proof for just 4% of RTE
problems.
• Translate to formal representation & apply automated reasoner
• Can succeed in restricted domains, but not in open-domain NLI!
Solution? Natural logic! ( natural deduction)
• Characterizes valid patterns of inference via surface forms
– precise, yet sidesteps difficulties of translating to FOL.
• A long history
– traditional logic: Aristotle’s syllogisms, scholastics, Leibniz, …
– modern natural logic begins with Lakoff (1970).
– van Benthem & Sánchez Valencia (1986-91): monotonicity calculus.
– Nairn et al. (2006): an account of implicatives & factives.
• Angeli & Manning (2009), McCartney & Manning (2014):
– extends monotonicity calculus to account for negation & exclusion.
– incorporates elements of Nairn et al.’s model of implicatives.
In other words
If I mutate a sentence in this specified way, do I preserve its truth?
Basic entailment lexical relations
The set of basic entailment relations
diagram symbo
l
name example
x y equivalence couch sofa
x ⊏ y forward entailment (strict)
crow ⊏ bird
x ⊐ y reverse entailment (strict)
European ⊐ French
x ^ y negation (exhaustive exclusion)
human ^ nonhuman
x | y alternation (non-exhaustive exclusion)
cat | dog
x y cover (exhaustive non-exclusion)
animal nonhuman
x # y independence hungry # hippo
Relations are defined for all semantic types: tiny ⊏ small, hover ⊏ fly,
kick ⊏ strike,
this morning ⊏ today, in Beijing ⊏ in China, everyone ⊏ someone, all ⊏
most ⊏ some
Relations are defined for all semantic types:
Small example
Entailment and semantic composition
• How the entailments of a compound expression depend on the entailments of its parts?
• Typically, semantic composition preserves entailment relations:
Projecting relations induced by lexical mutations
• Projection function. Two sentences differing only by a single lexical relation (downward).
Projection Examples
Join Table
Two projected relations for composition.
Proof by Alignment
PP
Linguistic analysis
• Tokenize & parse input sentences (future: & NER & coref & …)
• Identify items w/ special projectivity & determine scope
• Problem: PTB-style parse tree semantic structure!
Jimmy Dean refused to move without blue jeans
NNP NNP VBD TO VB IN JJ NNS
NP NP
VP
S
Solution: specify scope in PTB trees using Tregex [Levy & Andrew 06]
VP
VP
S
+ + + – – – + +
refuse
move
Jimmy Dean
without
jeans
blue
category: –/o implicatives examples: refuse, forbid, prohibit, …
scope: S complement pattern: __ > (/VB.*/ > VP $. S=arg)
projectivity: {:, ⊏:⊐, ⊐:⊏, ^:|, |:#, _:#, #:#}
P Gazprom today confirmed a two-fold increase in its gas price
for Georgia, beginning next Monday.
H Gazprom will double Georgia’s gas bill. yes
Alignment for NLI
• Linking corresponding words & phrases in two
sentences
• Most approaches to NLI depends on a facility for
alignment
Alignment example
unaligned content:
“deletions” from P
approximate match:
price ~ bill
phrase alignment:
two-fold increase ~ double
H (hypothesis)
P (
pre
mis
e)
Approaches to NLI alignment
• Alignment via semantic relatedness.
• W2V, GloVE, BERTH.
Phrase-based alignment representation
EQ(Gazprom1, Gazprom1)
INS(will2)
DEL(today2)
DEL(confirmed3)
DEL(a4)
SUB(two-fold5 increase6, double3)
DEL(in7)
DEL(its8)
…
Represent alignments by sequence of phrase edits: EQ, SUB,
DEL, INS
• One-to-one at phrase level (but many-to-many at token level)
• Avoids arbitrary alignment choices; can use phrase-based resources
Proof by Alignment
will depend on:
1. the lexical entailment relation generated by e: (e)
2. other properties of the context x in which e is applied
( , )
Lexical entailment relations
x e(x)
compound expression
atomic edit: DEL, INS, SUB
entailment relation
Example: suppose x is red car
If e is SUB(car, convertible), then (e) is ⊐
If e is DEL(red), then (e) is ⊏
Crucially, (e) depends solely on lexical items in e,
independent of context x.
But how are lexical entailment relations determined?
Lexical entailment relations: SUBs
(SUB(x, y)) = (x, y)
For open-class terms, use lexical resource (e.g. WordNet)
for synonyms: sofa couch, forbid prohibit
⊏ for hypo-/hypernyms: crow ⊏ bird, frigid ⊏ cold, soar ⊏ rise
| for antonyms and coordinate terms: hot | cold, cat | dog
or | for proper nouns: USA United States, JFK | FDR
# for most other pairs: hungry # hippo
Closed-class terms may require special handling
Quantifiers: all ⊏ some, some ^ no, no | all, at least 4 at most 6
Lexical entailment relations: DEL & INS
Generic (default) case: (DEL(•)) = ⊏, (INS(•)) = ⊐
– Examples: red car ⊏ car, sing ⊐ sing off-key
– Even quite long phrases: car parked outside since last week ⊏ car
– Applies to intersective modifiers, conjuncts, independent clauses, …
– This heuristic underlies most approaches to RTE! • Does P subsume H? Deletions OK; insertions penalized.
Special cases
– Negation: didn’t sleep ^ did sleep
– Implicatives & factives (e.g. refuse to, admit that): more complex
– Non-intersective adjectives: former spy | spy, alleged spy # spy
– Auxiliaries etc.: is sleeping sleeps, did sleep slept
Proof by Alignment
Example:
Common Sense Reasoning with Natural Logic
• Task: Given an utterance, and a large knowledge base of supporting facts. We want to know if the utterance is true or false.
Common Sense Reasoning for NLP
Common Sense Reasoning for Vision
Example search as graph search
Example search as graph search
Example search as graph search
Example search as graph search
Example search as graph search
Example search as graph search
Edges of the graph
Edge templates
“Soft” Natural Logic
• Likely (but not certain) inferences
– Each edge has a cost >=0
• Detail: Variation among edge instances of a template.
– WordNet:
– Nearest neighbours distance.
– Most other cases distance is 1.
– Let us call this edge distance f.
What natural logic can’t do
• Not a universal solution for NLI
• Many types of inference not amenable to natural logic – Paraphrase: Eve was let go Eve lost her job – Verb/frame alternation: he drained the oil ⊏ the oil drained – Relation extraction: Aho, a trader at UBS… ⊏ Aho works for
UBS – Common-sense reasoning: the sink overflowed ⊏ the floor
got wet – etc.
• Also, has a weaker proof theory than FOL – Can’t explain, e.g., de Morgan’s laws for quantifiers: Not all birds fly Some birds don’t fly
• Enables precise reasoning about semantic containment … • hypernymy & hyponymy in nouns, verbs, adjectives, adverbs
• containment between temporal & locative expressions
• quantifier containment
• adding & dropping of intersective modifiers, adjuncts
• … and semantic exclusion … • antonyms & coordinate terms: mutually exclusive nouns, adjectives
• mutually exclusive temporal & locative expressions
• negation, negative & restrictive quantifiers, verbs, adverbs, nouns
• … and implicatives and nonfactives
• Sidesteps myriad difficulties of full semantic interpretation
What natural logic can do