Comparing Brands with Sentiment Analysis

• Sentiment Analysis can be used for Social Brand Monitoring. Sentiment Analysis can be done on Social Media texts (twitter feeds / facebook posts etc.) to understand the polarity of the conversations by monitoring brands, competitors and industry. Customers keep on posting their opinions on different brands on social media, so mining the texts and extracting their opinions using sentiment analysis can be very useful.
• It can be used to analyze trends and gain insight into what customers want and how their likings change over time - it can be used to assist in analyzing consumer trends and tendencies.
• First we need to define keywords for the desired brand(s) / product(s). The add search filters (e.g. location, time) along with the search phrases. Use Twitter, Facebook, Amazon as sources of data.

• Can do sentiment / polarity analysis, emotion classification, topic modelling for useful actionable insights.

• The following two plots show the trends in the number of tweets and the overall sentiment scores for different smartphone brands in India, respectively.

• The following plot shows the wordcloud for different smartphone brands in India talked about on Twitter.

• The next plot shows the location specific sentiment scores (color-coded) for smartphone brands in India talked about on Twitter.