Fake News Detection Project (ML)

Fake News Detection 


Fake news

Fake news simply means - to incorporate information that leads people to the wrong path. Nowadays fake news is spreading a lot and people share this information without verifying it. 

Fake news detection is a severe yet challenging problem. The rapid rise of social networking platforms has not only yielded an enormous increase in information but has also stimulated the spread of fake news. Thus, the effect of fake news has been increasing, sometimes extending to the offline world and intimidating public safety. Given a large amount of web content, automatic fake news detection is a practical solution useful to all online content providers, to reduce the human time and effort to detect and prevent the spread of fake news. 


Detecting Fake News using Python


First, we will import libraries and the dataset. The dataset we’ll use for this project- we’ll call it news.csv which has a shape of 7796×4 (7796 rows and 4 columns).


Required packages:


pip install pandas

pip install numpy

pip install scikit-learn

           

You can download the dataset here.


Analyzing the data:

Next step is to analyze the data using functions like head(), describe(), count(), etc


        After cleaning the data


Data Cleaning:


Removing columns which have no significance in the prediction process like ID or S.No .    
                data=data.drop(axis=1,labels="Unnamed: 0")
                data=data.dropna()


Data Processing like extraction of features from the text


CountVectorizer tokenizes (breaking down a sentence or paragraph or any text into words) the text along with performing very basic preprocessing like removing the punctuation marks, converting all the words to lowercase, etc. It is also used to transform a given text into a vector based on the frequency (count) of each word that occurs in the entire text.

Splitting the dataset:

To train any machine learning model irrespective of what type of dataset is being used we have to split the dataset into training data and testing data.

           from sklearn.model_selection import train_test_split as tts

          x_train,x_test,y_train,y_test=tts(x_mat,y,test_size=0.2,random_state=10)

Here I have used the ‘train_test_split’ to split the data in 80:20 ratio i.e. 80% of the data will be used for training the model while 20% will be used for testing the model that is built out of it.

Selection of Algorithm

Decision Tree Algorithm:

A Decision Tree is a white box type of machine learning algorithm. It shares internal decision-making logic, which is not available in the black box type of algorithms such as Neural Network. Its training time is faster compared to the neural network algorithm. The time complexity of decision trees is a function of the number of records and number of attributes in the given data. The decision tree is a distribution-free or non-parametric method, which does not depend upon probability distribution assumptions. Decision trees can handle high-dimensional data with good accuracy.

Confusion Matrix

A Confusion matrix is an N x N matrix used for evaluating the performance of a classification model, where N is the number of target classes.

The accuracy of the prediction can be calculated by adding the diagonals of the matrix and dividing it by the sum of all the elements of the matrix.

        Get Project Source Code Here

References:

Comments

Post a Comment

Popular posts from this blog

ML Internship at Dlithe

Image
Internship at DLithe Being agile is the need of the hour. The world is in disruptive mode. It is looking for an agile workforce that is adaptable and produces quicker results. While academics and work provide required knowledge and experience, but individuals must bridge gaps in skills. In this digital world, DLithe is what helps you upgrade your skills and be competent to work in the agile world. Internship Journey The course started from the basics of the python programming language and then went on to machine learning which involved interesting projects like a chatbot, iris flower species identification,  message encryption-decryption, real estate price prediction, and many other real-life projects. We received a series of assignments every week that incrementally solved a real-world machine learning problem. Python Programming   The first few days of the internship were completely devoted to learning the basics of python which included topics like variables, ...
Read more

COVID Vaccination Data Analysis

Image
  COVID-19 Vaccination Data Analysis The Covid-19 pandemic is the most critical health disaster to hit the world. Predicting the trend of COVID-19 vaccination has become a challenge. Many healthcare professionals, statisticians, and researchers are tracking the spread of the virus in different parts of the world using a variety of approaches. The increase in a variety of vaccines developed by talented scientists and the vaccination process sparked curiosity to learn more about current immunization programs and a keen interest in finding meaningful information from the data. After looking up multiple websites, I found a few datasets to work with.  The complete code is available on  GitHub  along with the dataset . Data Collection First, we will import libraries and the dataset. The dataset we’ll use for this project- we’ll call it cowin-2.csv . Required packages:(pandas ,matplotlib,seaborn)           import  pandas  as  p...
Read more