An application of sentiment analysis: Analyzing Optima Italia customer's emails

Sentiment analysis it's a powerful instrument to study mood of a large amount of customers that often uses social network or blog to communicate to the world their opinion. Some study (Jansen et al., 2009) on the power of social network demonstrate that a lot of “posts” express opinions on a brand or a product. When we have to study sentiment in a market that is not accustomed to share opinions, like the Italian market of energy and telecommunication, or for a medium company without a stronger internet reputation, try to find the mood of the web it's very difficult. In our paper we want to try solving this problem applying the sentiment analysis methodology to the emails received by customers. However, mood of customer emails is usually not positive, thus it could be useful to adopt scoring techniques to weights customer's moods. This approach could be useful to extract more information about Customers Satisfaction. Additionally we want to use the mood of the emails of a customer like another variable in a forecast model to predict the abandon of customers.