Sentiment Analysis is another buzzword that we hear more and more often, but may not entirely understand what it does or how to use it. Sentiment Analysis application in business offers benefits to both companies and customers, such as improving products and services, identifying strengths and weaknesses of the competition, as well as targeted advertising.
Sentiment Analysis, also known as Sentiment Detection or Opinion Mining, is the classification of the sentiment polarity in a given text that may express people’s opinions, appraisals, attitudes, and emotions toward entities, individuals, issues, events, topics and their attributes. Put simply, it answers the question “how does the writer feel about this topic?” with the answer positive or negative, but may also include neutral.
In other words, Sentiment Analysis identifies words or patterns in language that carry positive, negative, or neutral sentiment. As humans, we interpret sentiment based on not only our knowledge of the language, but also social context. Computers are excellent at identifying linguistic patterns quickly, but they face accuracy challenges when the interpretation of sentiment is heavily dependent on context. For example, humans understand easily that the word cheap is positive when referring to price, but negative when referring to quality. There may also be cases where a word may simply state a fact in one domain, but carry sentiment in another. The word “red” in a hotel review (“the wall had a picture with red flowers”) simply states a fact, whereas in a technology review the word “red”, as in “my picture printed red”, would be negative. Consequently, computers need to be taught how to take into account this kind of context when classifying sentiment polarity. This is why many Sentiment Analysis tools have recently been taking the form of domain-specific or topic-specific Sentiment Analysis, to improve accuracy by limiting the scope to a known context.
In addition to polarity, we may also want to know how strongly negative or positive the opinion is by using degrees of sentiment. Some sentiment words or phrases carry stronger feelings. For example, “the movie was not that good” is less negative than “the movie was terrible”. Including the degrees of sentiment polarity in our analysis may offer better insights for decision making. Another aspect we may be interested in is the type of sentiment: joy, sadness, disgust, disappointment, surprise, fear, anticipation, trust, etc. For example, the sentence “I expected faster download speeds”, is not only negative, but also carries the sentiment of disappointment. Another example is the sentence “I’m concerned that the next Avengers film will not include my favorite Marvel superhero”, where the sentiment is negative and we can identify the emotion of fear. Similarly, in the positive sentence “I can’t wait to try out the new iPhone!”, we also have a sense of anticipation.
Where is Sentiment Analysis used?
With the appearance of social media, Sentiment Analysis has become a very popular way of mining people’s public opinions. Sentiment analysis has been used for a long time to analyze the answers to open-ended survey questions; however, this data is usually limited and requires considerable effort to collect. However, millions of people share their opinions freely on social media platforms and we can harness this public data to inform intelligent business decisions with Sentiment Analysis.
Company Product or Service Evaluation. A lot of companies use Sentiment Analysis for extracting and understanding their customers’ opinions on their products or services. Reviews on websites, forums, and social media provide helpful insights on the strengths and limitations of a company’s products and/or services. Those insights may be used to discover possible issues that need improvement, as domain-specific Sentiment Analysis may automatically categorize positive and negative reviews into relevant topics and sub-topics. Imagine the benefit to be gained from the quick summarization of millions of customer experiences, where customer feelings about topics like “product quality” can be tracked in real time. Let’s take a look at the following review:
A domain-specific Sentiment Analysis tool would identify that the customer has positive feelings toward certain aspects of the product, but not about its size.
Competitor Intelligence. While knowing the strengths and limitations of a company’s own products and services is helpful, it is also important to understand the status of its competitors. They may wish to incorporate features that customers appreciate in a competitor’s products or services, or avoid those that competitor customers do not enjoy. Below, the printer of a company is compared to a similar printer of a competitor company (Printer X). The company’s printer is preferred for its easy set up and interface, but one of the competitor’s printers has the added feature of a rear feed.
Market Intelligence. If an organization hosts products or services from other companies, it is useful to know what people enjoy and what needs improvement. This information can help a company decide whether to keep offering a product or service, as well as which products or services to promote. Similarly, if a company sells their products through other companies, they may analyze customer reviews to identify issues with the seller or new vendors to sell their product. The negative feelings associated with the brand in the following tweet indicate a possible decrease in sales; a company selling this brand may need to take this into account.
Another example is the tweet below:
The customer is dissatisfied with the selection of the products from a brand in a specific store; unless the brand addresses this issue with the store, it could have implications for their sales.
Advertising. A lot of the current advertising strategies are based on the inclusion of people’s reviews on products or services, also known as customer testimonials. Sentiment analysis allows companies to identify positive reviews they wish to showcase among the millions available, in order to gain the customers’ trust. Similarly, they may wish to showcase their superiority compared to their competitors, by identifying negative trends on the competitors’ products and services. For example, Gusto retweets positive customer reviews on their Twitter account, but also showcase some of them on their website as an advertising strategy.
 Liu, B., & Zhang, L. (2012). A Survey of Opinion Mining and Sentiment Analysis. In C .C. Aggarwal and C.X. Zhai (eds.), Mining Text Data, DOI 10.1007/978-1-4614-3223-4_13, © Springer Science+Business Media, LLC