Sentiment trend analysis
You may wonder why we are doing a sentiment-based analysis again, and the reason is a simple one: it is obvious analysis to do when the data is a large corpus of text. In our case, it is even more important as news and sentiment are closely related. If you can deduce the sentiment-based theme of a large corpus of news data, then it means that you have gained an important insight into what might be a long and tedious process of classifying each document manually. Simple at it may seem, it is one of the most coveted tools of any text data miner.
For our use case, we will do an interesting analysis. We will go through The Guardian's articles with a mention of Indian Prime Minister Narendra Modi and try to see how the sentiment trends about him have changed over the years.
Getting the data – not again
In the last section, we came to understand the building blocks of data gathering from normal web pages. We will build a strategy around that procedure to extract the necessary...
