Sentiment Analysis: Understanding Emotions and The BIG Role of Social Media!

In a world where we generate 2.5 quintillion bytes of data every day, sentiment analysis has become a key tool for making sense of that data. This has allowed companies to get key insights and automate all kinds of processes. A continuation of one of our previous articles is the concept of Sentiment Analysis.  Internet of Behavior builds on the concept of Sentiment Analysis. Taking the people’s opinion they display as text and analyzing it opens up a whole new revenue. From brands understanding how the general public is responding to a certain product and brand.

Sentiment Analysis and opinion mining

Examining the language people use to understand their thoughts and behaviors is not new in the sciences. Self-reported data is used to understand sentiment in consumers since the early 2000’s. However, the use of self-reported data has its limitations, as do most self-reported data. With sampling & surveys, most researchers rely heavily on consumers’ abilities to almost accurately recall their felt experiences, which may be difficult to verbalize and reconstruct. It’s in human nature to feel things differently at different times.

Today, social media platforms and their sub branches are popular transporters to study customer sentiment on a larger scale and within a natural setting due to the significant share of online conversations expressing consumers’ thoughts, feelings and opinions about products and brands. The sentiment analysis in textual and contextual content often relies on simple opinionated annotation tasks during which annotators must identify whether a sentence is positive, negative or neutral. Given the large volume of social media content, in this day and age, manual analysis is practically impossible. Since the recent advances in deep sentimental learning, the ability of algorithms to analyze text in proportion to the context has improved drastically. New use of advanced artificial intelligence and machine learning techniques can be an effective tool for doing in-depth research.

Opinion Mining

Text information can be broadly categorized into two main types: facts and opinions. Facts are objective expressions about something. Opinions are usually subjective expressions that describe people’s sentiments, appraisals, and feelings toward a subject or topic. Sentiment analysis, just as many other NLP problems, can be modeled as a classification problem where two sub-problems must be resolved:

• Classifying a sentence as subjective or objective, known as subjectivity classification.
• Classifying a sentence as expressing a positive, negative or neutral opinion, known as polarity classification.

In an opinion, the entity the text talks about can be an object, its components, its aspects, its attributes, or its features. It could also be a product, a service, an individual, an organization, an event, or a topic. MonkeyLearn segregates Sentiment Analysis into sub categories. You can read their full work here.

The Types of Sentiment Analysis

Fine-grained Sentiment Analysis

Sometimes, just saying something is positive or negative just isn’t enough. We need a more in-depth Analysis and hence we can further segment it into:

• Very positive
• Positive
• Neutral
• Negative
• Very negative

This is usually referred to as fine-grained sentiment analysis. You have definitely used this method before, E.g giving a review out of five star (which is on a scale of Very bad to Very Good).

Emotion detection

Emotion detection aims at detecting emotions like, happiness, frustration, anger, sadness, and the like. Many emotion detection systems resort to lexicons (i.e. lists of words and the emotions they convey) or complex machine learning algorithms.

One of the downsides of resorting to lexicon is that the technology yet is to master the use of sarcasm and modern lingo. For Example, today describing something as lit is a positive sentiment. 20 years later it was just meant as literally being on fire.

Aspect-based Sentiment Analysis

Usually, when analyzing the sentiment in subjects, for example products, you might be interested in not only whether people are talking with a positive, neutral, or negative polarity about the product, but also which particular aspects or features of the product people talk about. That’s what aspect-based sentiment analysis is about. In our previous example:

“The battery life of this camera is too short.”

The sentence is expressing a negative opinion about the camera, but more precisely, about the battery life, which is a particular feature of the camera.

Intent analysis

Intent is everything, what we say and we mean can be very different things. Hence the context is always important. If a stranger joins a conversation between you and your friend, they might be confused as well. This is because they would be unaware of the full context. Similarly, Machines sometimes also find it difficult to identify intents.

Importance of analyzing on social media

In this current era, not being a part of the global village created by the vast amount of social media applications, is considered bizarre. One way or another, the majority of the world’s population is surfing or expressing their emotions online, be it publicly or through a messaging app. Now let’s combine Sentiment analysis and why it is important to social media. Presence of social media in the current era is vital for any sort of business or a person etc. The image of an entity is based on their presence on social media. Based on the audiences, target accounts are chosen to display certain ads and the responses of the people are recorded. Taking the example of the world’s leading coffee company, Starbucks for example. The following data was analyzed using tweets:

With social media reports, you are tracking important metrics to check how well your content has performed. By adding data from social media sentiment analysis to your reports, you can do much more. As your brand grows, there will be more and more users talking about it online. You will start getting an overview on how customers are feeling about your brand. With that, you will also change your social media content strategies, as and when you are directly engaging with your customers.

Trends on Social Media

 Defining a trend from a business dictionary defines it as an activity of gradual change in a condition, output, or process, or an average or general tendency of a series of data points to move in a certain direction over time.

Recently, the word trending has become this phenomenon where every product, every content is trying to get into the trending page. The Social Media Platforms are rewarding the content that does well. And hoping on trend, as it seems like is the new norm.

Methods and Techniques to perform sentiment analysis

Performing sentiment analysis is a tricky situation, as for different types of texts we need different methods. Fundamentally, Sentiment analysis uses various NLP methods & algorithms. Out which we can divide into 3 categories:

• Rule based: These perform sentiment analysis based on the rules crafted manually.
• Automatic: Uses machine learning algorithms and learns from previous data.
• Hybrid: Tries to combine both in order to get the best of both worlds.

Rule-based Approaches are manually crafted to help identify the sentiment or subjectivity behind a task. Whilst accurate, this process takes long as rules have to be put in manually and often some words are forgotten. Some examples of Rule based approach include stemming, POS tagging and lexicons.

Contrary to rule-based approaches, automatic approach relies solely on machine learning algorithms. The model is initially allowed a training set from which it learns and later on the test set, performs its magic. Some popular Automatic learning algorithms include:

• Naïve Bayes: a large group of probability-based algorithms that use Bayes theorem to predict which category a text lies into.
• Linear Regression: A statistical algorithm used to predict a dependent variable given the independent variable is present.
• Deep learning: A connection of multiple neural networks that work together to mimic the human brain.

Final Thoughts

Being a concept that has risen to fame in the past couple of years, sentiment analysis rose when companies normalized surveys and feedback after customers bought their products. Fast forward to the modern era, social media has climbed up the ladder and people rather post their thoughts on their profile compared to telling the entity directly.

For More Articles Regarding all things AI and Data. Visit datafifty.com.