A learning dataset aimed at predicting the feedbacks for bengali blogs

Abstract

Since the innovation of blogging, social media began to explode its popularity. As huge amount of documents appearing in social media, there is an enormous opportunity for the analysis of such documents to find interesting relevance across this ocean of data. This work focuses on the analysis of documents appearing in blogs particularly in Bengali blogs. The key goal of this work is to prepare a machine learning dataset to analyse the data extracted from social media. Additionally the idea of a prediction model has been invoked which allows to forecast the number of feedbacks that a blog document is expected to receive in near future. For these experiments, blog documents were crawled from the Internet and processed to facilitate research.