Blog text analysis using topic modeling, named entity recognition and sentiment classifier combine

This paper describes our experimental work on computational analysis of socio-political blog data through a novel combine of sophisticated language processing and visualization techniques. We have designed an integrated framework by utilizing Topic Modeling, Entity Extraction and Sentiment Analysis; to draw sociologically relevant inferences from unstructured free form blogosphere data. The dataset comprised of more than 9290 blog posts on social-political events related to the Arab spring. We have tried to extract important inferences from the dataset; such as key themes, persons, places, organizations and overall sentiment orientation of the content around different entities in the texts. We have tried to validate the inferences obtained through manual and Google search trends statistics. The results obtained are quite relevant and demonstrate the usefulness of our approach for computational analysis of social media data.