Kaggle競賽題之——Sentiment Analysis on Movie Reviews
阿新 • • 發佈:2019-01-01
題目連結:https://www.kaggle.com/c/sentiment-analysis-on-movie-reviews
越來越喜歡iPython notebook了。以下所有工作都可以在一個頁面上完成,FireFox支援比Chrome要好。
資料集分為train.tsv和test.tsv。欄位以\t分隔,每一行有四個欄位:PhraseId,SentenceId,Phrase,Sentiment。
情感標識:
0 - negative
1 - somewhat negative
2 - neutral
3 - somewhat positive
4 - positive
import pandas as pd df = pd.read_csv('train.tsv',header=0,delimiter='\t') df.info() <class 'pandas.core.frame.DataFrame'> Int64Index: 156060 entries, 0 to 156059 Data columns (total 4 columns): PhraseId 156060 non-null int64 SentenceId 156060 non-null int64 Phrase 156060 non-null object Sentiment 156060 non-null int64 dtypes: int64(3), object(1)
df.head()Out[6]:
PhraseId | SentenceId | Phrase | Sentiment | |
---|---|---|---|---|
0 | 1 | 1 | A series of escapades demonstrating the adage ... | 1 |
1 | 2 | 1 | A series of escapades demonstrating the adage ... | 2 |
2 | 3 | 1 | A series | 2 |
3 | 4 | 1 | A | 2 |
4 | 5 | 1 | series | 2 |
直接用訓練集的前5行做分類準確性測試:In [13]: df.Sentiment.value_counts()/df.Sentiment.count() Out[13]: 2 0.509945 3 0.210989 1 0.174760 4 0.058990 0 0.045316 dtype: float64
分類準確率及結果:X_train = df['Phrase'] y_train = df['Sentiment'] import numpy as np from sklearn.feature_extraction.text import TfidfTransformer from sklearn.pipeline import Pipeline from sklearn.linear_model import LogisticRegression text_clf = Pipeline([('vect', CountVectorizer()), ('tfidf', TfidfTransformer()), ('clf', LogisticRegression()), ]) text_clf = text_clf.fit(X_train,y_train) X_test = df.head()['Phrase'] predicted = text_clf.predict(X_test) print np.mean(predicted == df.head()['Sentiment']) for phrase, sentiment in zip(X_test, predicted): print('%r => %s' % (phrase, sentiment))
0.8
'A series of escapades demonstrating the adage that what is good for the goose is also good for the gander , some of which occasionally amuses but none of which amounts to much of a story .' => 3
'A series of escapades demonstrating the adage that what is good for the goose' => 2
'A series' => 2
'A' => 2
'series' => 2
df.head()['Sentiment']
0 1
1 2
2 2
3 2
4 2
第一個分類錯誤。測試資料集:
test_df = pd.read_csv('test.tsv',header=0,delimiter='\t')
test_df.info()
<class 'pandas.core.frame.DataFrame'>
Int64Index: 66292 entries, 0 to 66291
Data columns (total 3 columns):
PhraseId 66292 non-null int64
SentenceId 66292 non-null int64
Phrase 66292 non-null object
dtypes: int64(2), object(1)
用訓練好的模型對測試資料集進行分類:
from numpy import savetxt
X_test = test_df['Phrase']
phraseIds = test_df['PhraseId']
predicted = text_clf.predict(X_test)
pred = [[index+156061,x] for index,x in enumerate(predicted)]
savetxt('../Submissions/lr_benchmark.csv',pred,delimiter=',',fmt='%d,%d',header='PhraseId,Sentiment',comments='')
提交結果:參考:http://scikit-learn.org/stable/tutorial/text_analytics/working_with_text_data.html