gkeswani92/twitter-sentiment-analyser

Twitter-Sentiment-Analyser

The basic idea behind sentiment mining and analysis is to get a general feedback of the masses regarding the product or service. Companies have been using sentiment analysis for to determine how well their product has been received in the market, what the people have liked and disliked etc. One potential side of sentiment analysis that has not receive a lot of attention yet is the use of sentiment mining in helping a lay man research about a product by creating an opinion based search engine.

Through this opinion based search engine, we aim to allow the user to analyse the overall sentiment on twitter about the product there are interested in. We aim to do this by giving the user the following options-

Live Tweet Analysis- This feature retrieves the most recent tweets from twitter about the query and performs sentiment analysis on those tweets. The returned tweets will be classified into positive or negative classes by the pre-trained Naïve Bayes model classifier. They will then be displayed along with graphical representations in the form of pie charts etc. so as to give the user an overall picture of products review

Feature based analysis- This feature is available only for mobile phones and retrieves the latest tweets from to rate the battery life, display, cost and camera on a scale of 1-10

Historical analysis of tweets- This feature is used to display graphs to show the sentiment about a particular brand over a period of time i.e. last week, 3 months, 6 months or one year as sometimes it is important to analyse the sentiment about a product or brand over a period of time rather than just in the present

Comparison of brands- This feature allows the user to compare the ratings of two products by retrieving the latest tweets

Location based analysis- This shows the user the classification of tweets into positive and negative and places these tweets in the form of green or red pins respectively on a world map as it allows them to visualise which part of the world the positive/negative feedback is coming from