Python 在训练了朴素贝叶斯文本分类算法后，如何预测单个文本文件的主题

我使用文本和训练数据训练并测试了朴素贝叶斯算法。现在我想预测单个文本文件的主题

这是我的密码

#importing test, train data
import sklearn.datasets as skd

categories = ['business', 'entertainment','local', 'sports', 'world']
sinhala_train = skd.load_files('Cleant data\stemmed_filtered_sinhala-set1', categories= categories, encoding= 'utf-8')
sinhala_test = skd.load_files('Cleant data\stemmed_filtered_sinhala-set2',categories= categories, encoding= 'utf-8')
name_file = "adaderana_67571.txt"
A = open(name_file, encoding='utf-8')
new_file = A.read()

from sklearn.feature_extraction.text import CountVectorizer
count_vectorization = CountVectorizer()
train_data_tf = count_vectorization.fit_transform(sinhala_train.data)
train_data_tf.shape

from sklearn.feature_extraction.text import TfidfTransformer
tfidf_trans = TfidfTransformer()
train_data_tfidf = tfidf_trans.fit_transform(train_data_tf)
train_data_tfidf.shape

from sklearn.naive_bayes import MultinomialNB
clf = MultinomialNB().fit(train_data_tfidf, sinhala_train.target)

test_data_tf = count_vectorization.transform(sinhala_test.data)
test_data_tfidf = tfidf_trans.fit_transform(test_data_tf)
predicted = clf.predict(test_data_tfidf)

from sklearn import metrics
from sklearn.metrics import accuracy_score
print("Accuracy of the model:", accuracy_score(sinhala_test.target, predicted))
print(metrics.classification_report(sinhala_test.target, predicted, target_names=sinhala_test.target_names)),
metrics.confusion_matrix(sinhala_test.target, predicted)

这是我的输出

Accuracy of the model: 0.864
               precision    recall  f1-score   support

     business       0.78      0.94      0.85       100
entertainment       0.95      0.86      0.90       100
        local       0.89      0.65      0.75       100
       sports       0.91      0.93      0.92       100
        world       0.83      0.94      0.88       100

    micro avg       0.86      0.86      0.86       500
    macro avg       0.87      0.86      0.86       500
 weighted avg       0.87      0.86      0.86       500

array([[94,  2,  4,  0,  0],
       [ 2, 86,  2,  4,  6],
       [19,  0, 65,  5, 11],
       [ 1,  3,  1, 93,  2],
       [ 5,  0,  1,  0, 94]], dtype=int64)

现在我想预测文本文件

new\u file

的主题

---

有人能帮我写代码来预测这个文本文件的主题吗。

我解决了我的问题。这是我用来预测主题的代码

docs_new1 = sinhala_test_1
docs_new = [docs_new1]
X_new_counts = count_vectorization.transform(docs_new)
X_new_tfidf = tfidf_trans.transform(X_new_counts)

predicted_topic = clf.predict(X_new_tfidf)

for doc, category in zip(docs_new, predicted_topic):
    topic = ( sinhala_train.target_names[category])
return topic