Abstract:
Since the technology evolves day by day, it is undeniable fact that the interaction between people and
technology has been changed dramatically. All these changes result in the huge amount of data and information
flow on the web. As it is said that data is the new oil in today’s century, it has become crucial to be able to
analyze and use this data appropriately. Social media is one of the crucial contributors to the issue as it is
producing terabytes of data every single day.
Natural Language Processing (NLP) is known as the well-known tool to recognize and interpret the
human language. Being as a branch of Artificial Intelligence, NLP helps to automate the relationship between
human and machine with the help of the structure of natural language. The goal of NLP is to understand the
human language and answer the questions accordingly by processing given human information.
Some of the most popular applications of NLP are virtual contacts like Siri, Alexa, and Google Assist.
The simplest way to visualize how NLP work with Siri is that it transforms human commands into numbers
for making it understandable for machines. Another application of NLP is chatbots of which job is to help
support teams solving issues by understanding human requests and giving responses accordingly. Other
applications of NLP such as spelling recommendations, automatic translations on social media, categorizing
receiving emails appropriately are also on trend. To sum up, we can say that NLP aims to make the humans’
life easier by creating interaction between humans and machine.
NLP applies mainly two techniques for establishing machine-human interaction: syntactic analysis
and semantic analysis. To be able to apply these analysis tasks, several sub-tasks are implemented.
Syntactic analysis consists of applying grammar rules to the text for identifying the structure of the
text, the organization between the words, and their relations:
• Tokenization – splitting a word or a sentence into smaller parts in order to make it more
understandable.
• PoS tagging – known as Part of Speech tagging, labelling tokenized text by its parts of speech.
• Lemmatization & Stemming – splitting the words into smallest meaningful forms.
• Removal of Stop words – removing frequently occurring words which has no contributions to
semantics.
Semantic analysis rather aims to identify the meaning of the text by analyzing each individual word. This
can be achieved by applying following sub-tasks:
• Word sense disambiguation – identifying the sense of word within the context.
• Relationship extraction – identifying the relations between entities of the given text.
One of the main business use cases of NLP that is applied widely is Sentiment and Emotion Analysis
which is also the main topic of this paper. Sentiment and Emotion Analysis identifies the sentiment and
emotion values of the input text and categorizes user’s opinion into 3 sentiment values (positive, negative,
neutral) and 8 Plutchik’s emotions (anger, anticipation, joy, trust, fear, surprise, sadness, and disgust) [28].
SEA (shortly Sentiment and Emotion Analysis) can be used for the analysis of social media comments,
organization reviews, online surveys, and customer service reviews. By this way, company leaders can clearly
see how their customers feel about their products and make appropriate decisions.
SEA can highly influence the organization’s productivity and quality as it helps to show the strong and
weak sides of the products based on customer review analysis.