Practical Decision Tree Implementation Of Machine Learning Techniques For Social Media Analysis

A large amount of data has been produced in engineering,biological and computational fields.With the passage of time,it becomes a difficult task to Analyzing a large amount of data.In every field of life,the information quantity is too large and exponentially increasing.Like,different social applications Facebook,Twitter,and Whats App etc.generates data of several terabytes every day in different forms like digital photos or videos,wall status or post,conversations etc.This is tricky task to analyze,manage and store the large datasets of social applications.Many machine learning techniques and algorithms are used for classification and analysis of datasets.Machine learning classification and regression algorithms include decision tree,na?ve bayes,logistic regression,SVM(Support Vector Machine),gradient boosted tree,random forest,generalized linear model,linear regression.Massive number of Tweets data is being generated during the Elections for different political parties like PPP,PML,PMLN and PTI in Pakistan.A lot of tools and platforms have been introduced with the advancement of new technologies and methods,but they do not perform better in terms of accuracy,scalability,and performance.The colossal volume of elections data involves computational as well as analytical challenges and requires better analysis in the perspective of the effective method.Over the time,machine learning techniques have been proved successful for data mining tasks like classification,prediction,and clustering.In the designing of data,ETL(Extraction,Transformation,and Loading)has the main problem today.The election tweets data analysis consists Classification phase according to their status.In this work,classification is performed using Anaconda Decision Tree(ADT)approach.Decision Tree(DT)classifier is used to classify the features i.e.PPP,PML,PMLN and PTI.For the Classification purpose,the proposed approach is used Python libraries with the help of Py Torch framework.The experimental results show that 98% accuracy is achieved by using Anaconda based Decision Tree Classifier.