我正在使用NLTK和SKlearn构建情感分析模型 . 该模型的准确性非常好 . 75％ . 模特受过7500个标签短片评论的培训，并在3000个标签短片评论中进行测试 .
但是我注意到它不承认否定（例如我不开心，我不是很好 - 模型将它们评为“积极”） . 我想我必须使用一些Bi-gram NLTK函数，但我不太确定如何添加当前代码 .
下面是我定义功能的部分代码
# open and read txt file with labelled dataset
short_pos = open("short_reviews_positive.txt","r").read()
short_neg = open("short_reviews_negative.txt","r").read()
all_words = []
documents = []
# tokinize and tag and allow only "adjective" "J"
allowed_word_types = ["J"]
for p in short_pos.split('\n'):
documents.append( (p, "pos") )
words = word_tokenize(p)
pos = nltk.pos_tag(words)
for w in pos:
if w[1][0] in allowed_word_types:
all_words.append(w[0].lower())
for p in short_neg.split('\n'):
documents.append( (p, "neg") )
words = word_tokenize(p)
pos = nltk.pos_tag(words)
for w in pos:
if w[1][0] in allowed_word_types:
all_words.append(w[0].lower())
# define features
all_words = nltk.FreqDist(all_words)
word_features = [w for (w, c) in all_words.most_common(5000)]
def find_features(document):
words = word_tokenize(document)
features = {}
for w in word_features:
features[w] = (w in words)
return features
featuresets = [(find_features(rev), category) for (rev, category) in documents]
random.shuffle(featuresets)
training_set = featuresets[:7500]
testing_set = featuresets[7500:]