What is Sentiment Analysis?
Sentiment essentially relates to feelings; attitudes, emotions and opinions. Sentiment Analysis refers to the practice of applying Natural Language Processing and Text Analysis techniques to identify and extract subjective information from a piece of text.
A person’s opinion or feelings are for the most part subjective and not facts. This means to accurately analyze an individual’s opinion or mood from a piece of text can be extremely difficult. With Sentiment Analysis from a text analytics point of view, we are essentially looking to get an understanding of the attitude of a writer with respect to a topic in a piece of text and its polarity; whether it’s positive, negative or neutral.
In recent years there has been a steady increase in interest from brands, companies and researchers in Sentiment Analysis and its application to business analytics. The business world today, as is the case in many data analytics streams, are looking for “business insight.”
So what “insights” are we talking about?
Well, in relation to sentiment analysis, we’re talking about insights into consumer behaviour, what our customers want, what our customers like and dislike about our products, what their buying signals are, what their decision process looks like etc…
As more and more content is created and shared online, through Social Channels, Blogs, Review Sites etc. we are becoming more and more vocal and open about our experiences online. In a recent study carried out by Zendesk it was noted that 45% of people share bad customer service experiences and 30% share good customer service experiences via social media. Which again, highlights the need and desire for businesses to mine this information to gain business insight from it has also increased.
Businesses are trying to unlock the hidden value of text in order to understand their customers’ opinions and needs and make better, more informed, business decisions. Traditionally businesses relied on surveys, workshops and focus groups to gain insight into their customers opinions and feelings, but today with modern technology we are able to harness the power of Machine Learning and Artificial Intelligence to extract meaning from text and dive into opinions of customers and see them outside of the often controlled environment of a survey.
What can we Analyze?
There is a wealth of information out there hidden in individuals' comments, emails, tweets, form submissions, reviews - the challenge is wrangling all of this info and extracting value from it. Below are some examples of detecting an individual’s opinion about a certain topic or product using tweets and reviews:
A tweet about Facebook’s messenger app:
“Literally ur facebook message app is useless, you only want it to increase profit. Please fix yourself. Its sad @facebook”
In this case we are going to detect the sentiment of the tweet in terms of polarity (positive or negative) and subjectivity (subjective or objective).
A product review:
“I really enjoyed using the Canon Ixus in Madrid. The Panasonic Lumix is a poor camera, but the Canon Ixus is really sleek. The Canon Ixus is much better than the Panasonic Lumix. All I want to do with a camera is point it and then just press the button. The Canon Ixus is perfect for that. You will soon get great photos with very little effort. I had previously returned a Panasonic Lumix because the pictures were not of the quality I expected. Spending the money on the Canon was a smart move. I would recommend the Ixus to anyone without hesitation. Two small issues to be pointed out: firstly, the battery life is not fantastic. Also, the zoom is not very powerful. If your subject is far away, you should not expect good results. Review made by John Irving from Florida.”
There is a couple of things you can do with this sort of information. We can act on them quite easily for example by responding in a timely manner to negative tweets and dealing with them head-on or showing your love for a positive review by saying thanks!
While it is somewhat interesting to look at each of these examples in isolation the true “business” value comes when you combine more and more tweets or reviews at scale. Looking at the bigger picture allows you to truly listen to the voice of your customer, identify trends, pinpoint problems and essentially extract more “business” value.
Interested in analyzing the sentiment of text, tweets, comments or reviews? Check out our Text Analysis API.
13 Oct, 2021
Why mainstream media coverage isn’t enough for media and risk intelligence companies
4 Min Read
29 Sep, 2021
News data as a signal in Supply Chain Risk monitoring
5 Min Read
28 Sep, 2021
How does Revolut use AYLIEN News API?
2 Min Read