You are on page 1of 4

CHAPTER 6

METHODOLOGY

6.1 Text Classification method – Review


Text classification is the act of dividing a set of input documents into two or more
classes where each document can be said to belong to one or multiple classes. Huge growth
of information flows and especially the explosive growth of Internet promoted growth of
automated text classification. Development of computer hardware provided enough
computing power to allow automated text classification to be used in practical applications.
Text classification is commonly used to handle spam emails, classify large text collections
into topical categories, manage knowledge and also to help Internet search engines.

6.2 Result and Analysis

We testify our proposed methods from the following aspects;

The effectiveness of general classifier built with training data and information from
web sources.

The influence of information gotten from different web sources for the
performance of general classifier.

The effectiveness of text classifier with derived features and tweets of given Naïve
Bayes.

In This work one dataset is used. This dataset is referred as number of people tweets the
comments of fake news. In order benchmark of dataset for the evaluation purpose, they have
collected the data, data formatted and cleaned.
Table

No Comments Neg/Pos
1 It is bad news chennal! Neg

2 It is bad news displaying in this chennal! Neg

3 I do not like this bad news Neg

4 I can't watch this news Neg

5 This is an amazing news Pos

6 This is good news Pos

7 I ain't feeling bad Neg

6.2 Work Process

The process of text classification as seen from the point of view of automatic text
classification systems can be clearly carried out in steps:

1. Collect the dataset of news


2. Collect both positive and negative comments
3. Use TextBlob to create your own text classification systems.
4. Import the textblob.classifiers and create some training and test data.
5. Create a new classifier by passing training data into the constructor for
a NaiveBayesClassifier.
6. Now classify the arbitrary text using the NaiveBayesClassifier.classify(text)
method.
7. Another way to classify strings of text is to use TextBlob objects
8. Call the classify() method on the blob.
9. Take advantage of TextBlob's sentence tokenization and classify each sentence
indvidually.
10. Now, let's check the accuracy on the test set.
11. We can also find the most informative features

Eg:
Accuracy=0.55555

Most Informative Features


# contains(In) = True neg : pos = 1.7 : 1.0
# contains(This) = True neg : pos = 1.7 : 1.0
# contains(An) = True neg : pos = 1.7 : 1.0
# contains(Content) = False neg : pos = 1.7 : 1.0
# contains(News) = True neg : pos = 1.7: 1.0

6.3 Work Terminology

In the present work, I have taken the news dataset regarding the comments on account of
their fake news problems. They are facing different account of problems such as heavy fake news.
Different comment their post on the problems that they are facing in their fake news detection.
They post both positive and negative type of comments of problems.

6.4 Text classification Implementation – Picturization

This paper will look through the state of the art in all these phases, take note of
methods and algorithms used for reducing computational complexity and improving the
precision of text classification process.

The overall work is picturized as below

Naive Bayes

Positive Negative
Comments Accuracy
Comments Predictio Value
n
News
Dataset

You might also like