Friday 13 May 2016

Sentiment analysis in the age of Digital Transformation

Sentiment Analysis is the process of determining whether a piece of information or service provided leads to positive, negative or neutral human feelings or opinions.

It is essentially, the process of extracting, identifying and characterizing the sentiments with the help of natural language processing, statistics, or machine learning methods and can be derived from various online mediums such as social media, review forums and blogs or as part of call center operations

Some of the sentiment prediction tools work just by looking at words in isolation, giving positive points for positive words and negative points for negative words and then summing up these points.

Sentiment analysis would help you to better plan your next course of marketing actions keeping in mind the existing tone.

Some of the questions that Sentiment analysis can help are:
  • Why don’t people like this new product?
  • Does this customer feedback look satisfied or dissatisfied?
  • What do consumers talk about my brand in online space?
  • Are they happy with the services?
  • What do our customers like about our competitors?
A typical process for sentiment analysis is as below:
  • Extract text or streaming data from various sources
  • Removing stop words, punctuation, slang words, special characters, numbers, extra white spaces
  • Stemming meaning putting word variations like "great", "greatly", "greatest", and "greater" all into one bucket
  • Converting text to lower case
  • Categorization using dictionaries (taxonomies) according to the line of business
  • Identifying one word, two words or three-words combinations for more accuracy
  • Classify the words into positive, negative or neutral categories
  • Generate the word clouds
  • Provide further reports on most negative sentiments to be actioned by business

Today there are several products available to do sentiment analysis using Natural Language Processing (NLP) & Machine Learning.

They go across all the social media conversations like blogs, news, forums, videos, tweets, reviews, images, Facebook etc and collect the data streams for further analysis.

Several machine learning algorithms like decision trees, naive Bayes classification are used for classification.

NLP provides the ability to read and understand, as well as derive meaning from the languages that humans speak, and it is part of Artificial Intelligence.

Nestle, via their Digital Acceleration Team, tracks the sentiments of their 2000+ brands to know what their customers think and to deliver products that they want and to prevent crisis from happening.

Coca-Cola, the brand that built its marketing message around happiness and sharing, has built vending machines which set the price of a can based on how positive your tweets are. 

Consumers are always on their smartphones leaving the trails of their feelings in the digital world.

In the age of Digital Transformation, Sentiment analysis is all about helping companies gain better insights into their customers, and helping them to bridge the gap between insight and action.

1 comment:

  1. Sentiment analytics cannot be relevant today without using NLP

    ReplyDelete

360TotalSecurity WW