sentiment.py 3.5 KB
Newer Older
W
wen-bo-yang 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
# /usr/bin/env python
# -*- coding:utf-8 -*-

# Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserved
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""
The script fetch and preprocess movie_reviews data set

that provided by NLTK
"""

W
wen-bo-yang 已提交
23 24
import paddle.v2.dataset.common as common
import collections
W
wen-bo-yang 已提交
25 26
import nltk
import numpy as np
27
from itertools import chain
W
wen-bo-yang 已提交
28 29
from nltk.corpus import movie_reviews

W
wen-bo-yang 已提交
30
__all__ = ['train', 'test', 'get_word_dict']
W
wen-bo-yang 已提交
31 32 33 34 35 36 37 38 39 40
NUM_TRAINING_INSTANCES = 1600
NUM_TOTAL_INSTANCES = 2000


def download_data_if_not_yet():
    """
    Download the data set, if the data set is not download.
    """
    try:
        # make sure that nltk can find the data
W
wen-bo-yang 已提交
41 42
        if common.DATA_HOME not in nltk.data.path:
            nltk.data.path.append(common.DATA_HOME)
W
wen-bo-yang 已提交
43 44 45
        movie_reviews.categories()
    except LookupError:
        print "Downloading movie_reviews data set, please wait....."
W
wen-bo-yang 已提交
46
        nltk.download('movie_reviews', download_dir=common.DATA_HOME)
47 48
        print "Download data set success....."
        print "Path is " + nltk.data.find('corpora/movie_reviews').path
W
wen-bo-yang 已提交
49 50 51 52 53 54 55 56 57


def get_word_dict():
    """
    Sorted the words by the frequency of words which occur in sample
    :return:
        words_freq_sorted
    """
    words_freq_sorted = list()
W
wen-bo-yang 已提交
58
    word_freq_dict = collections.defaultdict(int)
W
wen-bo-yang 已提交
59
    download_data_if_not_yet()
W
wen-bo-yang 已提交
60 61 62 63 64 65

    for category in movie_reviews.categories():
        for field in movie_reviews.fileids(category):
            for words in movie_reviews.words(field):
                word_freq_dict[words] += 1
    words_sort_list = word_freq_dict.items()
W
wen-bo-yang 已提交
66 67
    words_sort_list.sort(cmp=lambda a, b: b[1] - a[1])
    for index, word in enumerate(words_sort_list):
W
wen-bo-yang 已提交
68
        words_freq_sorted.append((word[0], index))
W
wen-bo-yang 已提交
69 70 71
    return words_freq_sorted


72 73 74 75 76 77 78 79 80 81 82 83 84
def sort_files():
    """
    Sorted the sample for cross reading the sample
    :return:
        files_list
    """
    files_list = list()
    neg_file_list = movie_reviews.fileids('neg')
    pos_file_list = movie_reviews.fileids('pos')
    files_list = list(chain.from_iterable(zip(neg_file_list, pos_file_list)))
    return files_list


W
wen-bo-yang 已提交
85 86 87 88 89 90
def load_sentiment_data():
    """
    Load the data set
    :return:
        data_set
    """
91
    data_set = list()
W
wen-bo-yang 已提交
92
    download_data_if_not_yet()
93 94 95 96 97 98 99
    words_ids = dict(get_word_dict())
    for sample_file in sort_files():
        words_list = list()
        category = 0 if 'neg' in sample_file else 1
        for word in movie_reviews.words(sample_file):
            words_list.append(words_ids[word.lower()])
        data_set.append((words_list, category))
W
wen-bo-yang 已提交
100 101 102 103 104 105 106 107 108 109
    return data_set


def reader_creator(data):
    """
    Reader creator, it format data set to numpy
    :param data:
        train data set or test data set
    """
    for each in data:
W
wen-bo-yang 已提交
110
        yield each[0], each[1]
W
wen-bo-yang 已提交
111 112 113 114 115 116


def train():
    """
    Default train set reader creator
    """
W
wen-bo-yang 已提交
117
    data_set = load_sentiment_data()
W
wen-bo-yang 已提交
118 119 120 121 122 123 124
    return reader_creator(data_set[0:NUM_TRAINING_INSTANCES])


def test():
    """
    Default test set reader creator
    """
W
wen-bo-yang 已提交
125
    data_set = load_sentiment_data()
W
wen-bo-yang 已提交
126
    return reader_creator(data_set[NUM_TRAINING_INSTANCES:])