[Language Processing and Python] 10.5 paragraph semantic Layer

A paragraph is a sequence of sentences.

Paragraph Representation Theory

The Quantization Standard Method in the first-order logic is limited to a single sentence, but some quantifiers can be expanded to more than two sentences.

See the following example:

(54&own(Angus, x)&bite(x, Irene))

The objective of Discourse RepresentationTheory (DRT) is to provide a way to process this and other semantic phenomena that seem to be the features of a paragraph.

DRS (discourse representation structure, DRS) Section representation structure

:

>>> Drs1 = dp. parse (>>>

You can view the visual effect:

>>>drs1.draw()

& Dog (y) & own (x, y ))

The DRT expression has the DRS join operator, which is expressed as +.

>>>drs2 = dp.parse(>>>+>>>

One DRS is embedded into another DRS. This is the way in which the full name quantifiers are processed.

 

>>>drs3 = dp.parse(>>>-> exists y.(ankle(y) &bite(x,y)))

If DRS contains conditions in the PRO (x) format, replace the resolve_anaphora () method with x = [...] format conditions, where [...] is a possible first-line word linked list.

>>>drs4 = dp.parse(>>>drs5 = dp.parse(>>>drs6 = drs4 +>>>>>>= [x,y,z]), irene(z), bite(u,z)])

The existing mechanism of DRS processing and processing λ abstraction is compatible.

Det[NUM=sg,SEM=<\P Q.([x],[])+ P(x)+ Q(x)>]-> =sg,SEM=<\P Q.exists x.(P(x)&Q(x))>]-> 

For example, a dog:

(NP[NUM=, SEM=<\Q.(([x],[dog(x)])+ Q(x))>, SEM=<\PQ.((([x],[])+ P(x))+ Q(x))>=, SEM=<\x.([],[dog(x)])>=, SEM=<\x.([],[dog(x)])>]dog)))))

We can use the DRT Parsing Method to parse sentences:

 

>>> nltk >>>parser= load_parser(, logic_parser=>>>trees = parser.nbest_parse(>>> trees[0].node[

Paragraph Processing

A paragraph is a sentence sequence, s1, s2, s3... The paragraph line is the sequential s1-ri of reading ,... Sn-f.

>>>dt =nltk.DiscourseTester([,>>>-r0: exists x.(student(x)&-r0: all x.(student(x)-> person(x))

We can add sentences and delete sentences at any time. Setting consistchk = True can check the reading sequence that can be received to check the consistency of the module:

>>>dt.add_sentence(,consistchk=, , -r0: exists x.(student(x)&-r0: all x.(student(x) ->-r0: -exists x.(person(x)&>>>dt.retract_sentence(,verbose= a person

We use informchk = True to check whether the new sentence.

 

>>>dt.add_sentence(,informchk=under reading