Skip to content
GitLab
Explore
Sign in
Primary navigation
Search or go to…
Project
S
Statistical_Grammar_Checker
Manage
Activity
Members
Labels
Plan
Issues
Issue boards
Milestones
Iterations
Wiki
Requirements
Code
Merge requests
Repository
Branches
Commits
Tags
Repository graph
Compare revisions
Snippets
Locked files
Build
Pipelines
Jobs
Pipeline schedules
Test cases
Artifacts
Deploy
Releases
Package registry
Container registry
Model registry
Operate
Environments
Terraform modules
Monitor
Incidents
Analyze
Value stream analytics
Contributor analytics
CI/CD analytics
Repository analytics
Code review analytics
Issue analytics
Insights
Model experiments
Help
Help
Support
GitLab documentation
Compare GitLab plans
Community forum
Contribute to GitLab
Provide feedback
Keyboard shortcuts
?
Snippets
Groups
Projects
Show more breadcrumbs
NLP_WS_2021
Statistical_Grammar_Checker
Commits
1811e34b
Commit
1811e34b
authored
3 years ago
by
jmzk96
Browse files
Options
Downloads
Patches
Plain Diff
push after meeting 25.11.2021
parent
2a535025
Branches
Branches containing commit
No related tags found
No related merge requests found
Changes
1
Hide whitespace changes
Inline
Side-by-side
Showing
1 changed file
src/grammar_checker.py
+44
-46
44 additions, 46 deletions
src/grammar_checker.py
with
44 additions
and
46 deletions
src/grammar_checker.py
+
44
−
46
View file @
1811e34b
...
...
@@ -21,6 +21,42 @@ class GrammarChecker:
with
open
(
f
"
../data/corpus_n=
{
n
}
.json
"
,
"
r
"
)
as
infile
:
self
.
corpus
=
json
.
load
(
infile
)
def
get_probs_of_gram
(
self
,
gram
:
Union
[
list
,
tuple
])
->
float
:
# Get first character of toekn/word in gram
first_char
=
gram
[
0
][
0
]
# Get length of gram -1 for preceeding tokens
length_gram
=
len
(
gram
)
-
1
# Get first n-1 tokens in gram
first_n_1_token
=
gram
[:
length_gram
]
# Search corpus for first character
search_corpus
=
self
.
corpus
[
first_char
]
#Initiate count for total of grams with preceding tokens and
#also count for total grams with exactly the same gram
total_value
=
0
conditional_value
=
0
for
key
,
value
in
search_corpus
.
items
():
# split words into list of tokens//grams in corpus
splitted
=
key
.
split
()
if
splitted
[:
length_gram
]
==
list
(
first_n_1_token
):
total_value
+=
value
if
splitted
==
list
(
gram
):
conditional_value
+=
value
try
:
prob
=
conditional_value
/
total_value
if
prob
==
0.0
:
prob
=
1
/
sys
.
maxsize
except
ZeroDivisionError
:
prob
=
1
/
sys
.
maxsize
return
prob
def
get_chained_probability
(
self
,
list_of_ngrams
:
list
)
->
list
:
list_of_probs
=
[
self
.
get_probs_of_gram
(
i
)
for
i
in
list_of_ngrams
]
# list_of_cum_probs = [get_probs_of_gram(i,dictionary) for i in split_sliding_ngram(input,n)[:position]]
list_of_cum
=
[
list_of_probs
[
0
]]
for
i
in
range
(
1
,
len
(
list_of_probs
)):
list_of_cum
.
append
(
list_of_cum
[
-
1
]
*
list_of_probs
[
i
])
return
list_of_cum
def
check
(
self
,
sentence
):
"""
checks a sentence for errors and recursively corrects the first one
"""
# add padding
...
...
@@ -28,9 +64,9 @@ class GrammarChecker:
sentence
=
self
.
start_tag
*
(
self
.
n
-
1
)
+
sentence
+
self
.
end_tag
*
(
self
.
n
-
1
)
# create n_grams list
n_grams
=
list
(
nltk
.
ngrams
(
self
.
tokenizer
.
tokenize
(
sentence
),
self
.
n
))
# find
error
i_error
=
self
.
find_
index_of_er
ro
r
(
n_grams
)
print
(
f
"
erroneous n-gram:
\n
{
n_grams
[
i_error
]
}
"
)
# find
first error and it's correction
i_error
=
self
.
find_
gram_lowest_p
ro
b
(
n_grams
)
print
(
n_grams
[
i_error
])
# if error detected
# if i_error:
# # find correction
...
...
@@ -67,48 +103,10 @@ class GrammarChecker:
# return i
# return None
def
get_prob_of_gram
(
self
,
n_gram
):
"""
calculates probability of n_gram
"""
# smallest possible positive float (1e-324 == 0.0)
float_min
=
1e-323
# get first character n_gram and n_gram without last word
first_char
=
n_gram
[
0
][
0
]
all_but_last_tokens
=
n_gram
[:
-
1
]
# if n_gram isn't in corpus
if
n_gram
not
in
self
.
corpus
[
first_char
]:
# return smallest possible positive float
return
float_min
# get n_gram occurences and total occurences starting with the same n-1 words
n_gram_occurrences
=
self
.
corpus
[
first_char
][
n_gram
]
total_value
=
0
for
key
,
value
in
self
.
corpus
[
first_char
].
items
():
# split key string into list of tokens
splitted
=
key
.
split
(
"
"
)
# if first n-1 words are the same as of n_gram
if
splitted
[:
-
1
]
==
list
(
all_but_last_tokens
):
# add occurences to total number of occurrences
total_value
+=
value
# calculate n_gram probability
prob
=
conditional_value
/
total_value
# return it if it's not 0, else return smallest possible positive float
return
prob
if
prob
!=
0.0
else
smallest_possible_positive_float
def
get_chained_probabilities
(
self
,
n_grams
):
"""
get list of cumulative markov chains for n_grams
"""
probs
=
[
self
.
get_prob_of_gram
(
n_gram
)
for
n_gram
in
n_grams
]
print
(
f
"
list of probabilities:
\n
{
probs
}
\n
"
)
chained_probs
=
[
probs
[
0
]]
for
i
in
range
(
1
,
len
(
list_of_probs
)):
chained_probs
.
append
(
chained_probs
[
-
1
]
*
probs
[
i
])
print
(
f
"
list of chained probabilities:
\n
{
chained_probs
}
\n
"
)
return
chained_probs
def
find_index_of_error
(
self
,
n_grams
):
"""
finds index of greatest error in n_grams
"""
chained_probs
=
self
.
get_chained_probabilities
(
n_grams
)
# calculate differences between values in chained_probs
diff_list
=
[
abs
(
j
-
i
)
for
i
,
j
in
zip
(
chained_probs
[:
-
1
],
chained_probs
[
1
:])]
print
(
"
list of differences:
\n
{chained_probs}
\n
"
)
def
find_gram_lowest_prob
(
self
,
ngrams
:
list
):
probs_list
=
self
.
get_chained_probability
(
ngrams
)
diff_list
=
[
abs
(
j
-
i
)
for
i
,
j
in
zip
(
probs_list
[:
-
1
],
probs_list
[
1
:])]
print
(
probs_list
,
"
\n
"
,
diff_list
)
return
diff_list
.
index
(
max
(
diff_list
))
+
1
def
find_correction
(
self
,
bad_n_gram
):
...
...
@@ -126,7 +124,7 @@ class GrammarChecker:
if
__name__
==
"
__main__
"
:
text
=
"
Note that in all contour time-integral we essentially integrate
over
_MATH_.
"
text
=
"
Note that in all contour time-integral
s
we essentially integrate
s
_MATH_.
"
# text = "Optimal filters was categorized to recursive and batch filters."
grammar_checker
=
GrammarChecker
(
3
)
...
...
This diff is collapsed.
Click to expand it.
Preview
0%
Loading
Try again
or
attach a new file
.
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Save comment
Cancel
Please
register
or
sign in
to comment