SENTIMENT ANALYSIS OR OPINION MINING

With the proliferation and standardization of on-line expression forms, such as those that allow users to comment, like, reblog, retweet, repin, rate, and review, the use of big data processing techniques, such as sentiment analysis, accelerates as user-content-generated continue to grow throughout the social media networks. Due to the amount, rate, variety of data being collected, the challenge of understanding and evaluating human behaviours becomes of greater complexity and importance. In fact data means nothing without data analysis; to become meaningful, data must be contextualised.

 

Sentiment analysis or opinion mining is a language-processing technique which examines a body of text to categorise opinion trends with regards to a specific subject or product. By analysing sites such as Amazon, Twitter or Facebook, it is possible for corporations, marketers, and governments to measure the overall attitude (judgment, evaluation or affective state) of the public about a particular product, service or idea. This makes possible to identify different perceptions in different demographics niches with respect to a product launch, a new service or a political campaign, and therefore evaluate the success or not of its implementation.

 

The use of data analysis to process and interpret big data opens up many possibilities, but also involves numerous risks. The algorithms used in these processes can oversimplify the complexity of an issue leaving out of the equation cultural values and diversity considerations.