cleantweet 2.1.0

a python library to clean textual data fetched from API's

CleanTweet version 2.1.0

CleanTweet is a python library created for cleaning textual data fetched from API. In the second version of the cleantweet library, there are 6 new classes that have been introduced. They include the DiagramTweet

PredictionTweet, SaveTweet, SentimentTweet, StatsTweet, and SyntheticTweet. Each of these classes have different methods specific

to each class and most of them inherit the clean() method from the CleanTweet class which was the only class present in the version

0.1.1.

You can find more detailed documentation about the various classes, their implementation and their various methods (close to 50) in

cleantweet's permanent home: https://cleantweet.org/

Installation

pip install cleantweet

Get Started

How to clean your Twitter Object:

Example 1: If the text file containing the Twitter/X JSON Data is in the same directory as project files.

!pip install cleantweet

import cleantweet as clt

Instantiate the CleanTwitter Object

data = clt.CleanTweet('sample_text.txt')

Call the clean method

data = data.clean())

Print the data object

print(data)

Various Implementations for the new classes

Draw a Wordcloud from your textual data

import cleantweet as clt

diagram = clt.DiagramTweet('sample_text.txt')

diagram.word_cloud()

Amount of Plural Proper Nouns in your text

import cleantweet as clt

text = clt.StatsTweet('sample_text.txt')

text = text.amount_of_proper_nouns(plural=True)

print(text)

Sentiment Analysis

import cleantweet as clt

sentiments = clt.SentimentTweet('sample_text.txt')

sentiments = sentiments.vader()

print(sentiments)

Synthetic Data

import cleantweet as clt

synths = clt.SyntheticTweet('sample_text.txt')

synths = synths.rule_based()

print(synths)

Update

For version 2.1.0, a new class was created called the ClassificationTweet class; the implementation is the same with the

other classes - create an instance of the class while passing the document as a .txt file, then using that class

instance to have access to the different methods in the class.

New methods have been added to the DiagramTweet class like pie_chart() and the lexicon_chart().

InferenceTweet class has been enhanced with frequency-based and TextRank summarization methods.

Named Entity Recognition (NER) has been implemented in the StatsTweet class for extracting named entities from text.