【NLP】使用 NLTK 解析与生成上下文无关文法（Context-free Grammar）

参考 链接

Context-free grammars with NLTK

NLTK定义了许多对使用自然语言很有用的类。在这里，我们看看如何使用NLTK中的上下文无关语法。

import nltk

groucho_grammar = nltk.CFG.fromstring("""
S -> NP VP
PP -> P NP
NP -> Det N | Det N PP | 'I'
VP -> V NP | VP PP
Det -> 'an' | 'my'
N -> 'elephant' | 'pajamas'
V -> 'shot'
P -> 'in'
""")

groucho_grammar

<Grammar with 13 productions>

type(groucho_grammar)

nltk.grammar.CFG

groucho_grammar.start()

S

groucho_grammar.productions()

[S -> NP VP,
 PP -> P NP,
 NP -> Det N,
 NP -> Det N PP,
 NP -> 'I',
 VP -> V NP,
 VP -> VP PP,
 Det -> 'an',
 Det -> 'my',
 N -> 'elephant',
 N -> 'pajamas',
 V -> 'shot',
 P -> 'in']

from nltk.grammar import *
groucho_grammar.productions(lhs=Nonterminal("NP"))

[NP -> Det N, NP -> Det N PP, NP -> ‘I’]

groucho_grammar.productions(rhs=Nonterminal("Det"))

[NP -> Det N, NP -> Det N PP]

pp = groucho_grammar.productions(rhs=Nonterminal("Det"))
pp[0]

NP -> Det N

pp[0].lhs()

NP

pp[0].rhs()

(Det, N)

解析（Parsing）

NLTK附带了用于cfg的预先实现的解析器。用CFG解析一个句子会返回一个解析树列表。我们可以查看它们的字符串表示形式，也可以用图形化的方式绘制树

sent = ['I', 'shot', 'an', 'elephant', 'in', 'my', 'pajamas']
parser = nltk.ChartParser(groucho_grammar)
trees = list(parser.parse(sent))

print(trees[0])

(S
  (NP I)
  (VP
    (VP (V shot) (NP (Det an) (N elephant)))
    (PP (P in) (NP (Det my) (N pajamas)))))

type(trees[0])

nltk.tree.Tree

trees[0]

trees[1]

生成

from nltk.parse.generate import generate
for sentence in generate(groucho_grammar, depth=5):
    print(' '.join(sentence))

an elephant shot an elephant
an elephant shot an pajamas
an elephant shot my elephant
an elephant shot my pajamas
an elephant shot I
an elephant shot I in I
an pajamas shot an elephant
an pajamas shot an pajamas
an pajamas shot my elephant
an pajamas shot my pajamas
an pajamas shot I
an pajamas shot I in I
my elephant shot an elephant
my elephant shot an pajamas
my elephant shot my elephant
my elephant shot my pajamas
my elephant shot I
my elephant shot I in I
my pajamas shot an elephant
my pajamas shot an pajamas
my pajamas shot my elephant
my pajamas shot my pajamas
my pajamas shot I
my pajamas shot I in I
an elephant in I shot an elephant
…
I shot my elephant
I shot my pajamas
I shot I
I shot I in I