Beispiel #1
0
import hmm_utils
import cmd_utils
import parsers


def round_to(n, precission):
    correction = 0.5 if n >= 0 else -0.5
    return int(n/precission+correction)*precission


counts = hmm_utils.get_transition_counts()

# Flags that note that incode should be looked for in STDIN instead of
# in a test essay file
grade_directory = cmd_utils.cmd_arg('--dir', None)
final_score_stdin = cmd_utils.cmd_flag('--final-score', None)
parse_stdin = cmd_utils.cmd_flag('--parse', None)
score_stdin = cmd_utils.cmd_flag('--score', None)
pronoun_stdin = cmd_utils.cmd_flag('--pronoun', None)
topic_stdin = cmd_utils.cmd_flag('--topic', None)
syntactic_formation_stdin = cmd_utils.cmd_flag('--syn-formation', None)
agreement_stdin = cmd_utils.cmd_flag('--agree', None)
sentence_parse_stdin = cmd_utils.cmd_flag('--sen-token', None)
word_order_parse_stdin = cmd_utils.cmd_flag('--word-order', None)


transition_count = cmd_utils.cmd_arg('--count', None)
transition_prob = cmd_utils.cmd_arg('--prob', None)


if grade_directory:
Beispiel #2
0
import hmm_utils
import cmd_utils
import parsers


def round_to(n, precission):
    correction = 0.5 if n >= 0 else -0.5
    return int(n / precission + correction) * precission


counts = hmm_utils.get_transition_counts()

# Flags that note that incode should be looked for in STDIN instead of
# in a test essay file
grade_directory = cmd_utils.cmd_arg('--dir', None)
final_score_stdin = cmd_utils.cmd_flag('--final-score', None)
parse_stdin = cmd_utils.cmd_flag('--parse', None)
score_stdin = cmd_utils.cmd_flag('--score', None)
pronoun_stdin = cmd_utils.cmd_flag('--pronoun', None)
topic_stdin = cmd_utils.cmd_flag('--topic', None)
syntactic_formation_stdin = cmd_utils.cmd_flag('--syn-formation', None)
agreement_stdin = cmd_utils.cmd_flag('--agree', None)
sentence_parse_stdin = cmd_utils.cmd_flag('--sen-token', None)
word_order_parse_stdin = cmd_utils.cmd_flag('--word-order', None)

transition_count = cmd_utils.cmd_arg('--count', None)
transition_prob = cmd_utils.cmd_arg('--prob', None)

if grade_directory:
    import os
    import grade_utils
    (2, 2, 2, 1,),
    (1, 2, 1, 2,),
    (3, 3, 2, 2,),
    (2, 3,),
    (2, 2, 1,),  # 15
    (1, 1, 1, 3, 1),
    (2, 1,),
    (2, 2, 1,),
    (2, 3,),
    (2, 2, 2, 2),  # 20
)

counts = hmm_utils.get_transition_counts()
essay_index = int(cmd_utils.cmd_arg('--essay', 0)) - 1
line_index = int(cmd_utils.cmd_arg('--line', -1))
use_stdin = cmd_utils.cmd_flag('--stdin')


# How much to prefer long answers over shorter onces
weight = .0001

invalid_boundary_tags = ('IN', 'CC', 'SINV', 'RP', 'TO')
pers_pro_tags = ('PRP', 'PRP$')
start_pers_pro_weight = 1000


def _possible_sentences_in_line(line, min_sentence_len=3):
    # The simplest thing here is to defer to the paper.  If it looks like they've
    # added punctuation already, lets just use that
    has_abbr = sum([1 if len(word) > 0 and word[-1] == "." and word.count(".") > 1 else 0 for word in line.split(" ")])
    if not has_abbr and line.count(". ") > 0:
Beispiel #4
0
    (
        2,
        2,
        1,
    ),
    (
        2,
        3,
    ),
    (2, 2, 2, 2),  # 20
)

counts = hmm_utils.get_transition_counts()
essay_index = int(cmd_utils.cmd_arg('--essay', 0)) - 1
line_index = int(cmd_utils.cmd_arg('--line', -1))
use_stdin = cmd_utils.cmd_flag('--stdin')

# How much to prefer long answers over shorter onces
weight = .0001

invalid_boundary_tags = ('IN', 'CC', 'SINV', 'RP', 'TO')
pers_pro_tags = ('PRP', 'PRP$')
start_pers_pro_weight = 1000


def _possible_sentences_in_line(line, min_sentence_len=3):
    # The simplest thing here is to defer to the paper.  If it looks like they've
    # added punctuation already, lets just use that
    has_abbr = sum([
        1 if len(word) > 0 and word[-1] == "." and word.count(".") > 1 else 0
        for word in line.split(" ")