Article Preview
TopIntroduction
With the advent of technology, social networking websites have become a popular medium for the users to express their sentiments, thoughts and opinions on various topics of interest which may either be any event or be any person or product. Social media channels have also got a foothold as far as the discussions amongst people on various trendy topics are concerned. People throughout the world can freely express their ideas with this type of platform to convey their outlook on any topic on earth. This has also become a platform for interacting with others. With this, the volume of data generated has become enormous. Users post more than 340 million tweets and 1.6 billion search queries every day (Bharti, Vachha, Pradhan, Babu & Jena, 2016). With this huge amount of data being generated, a number of challenges crop up. These challenged include accessing, storing and processing of data, verification of data sources, dealing with misinformation. There is another challenge in dealing with unstructured data as on an average, almost 80% of generated data are of this form.
In social media platforms, user interactions may vary from a simple text messaging to interacting through different multimedia channels such as images and videos. Users are often engaged in providing real-time feedback about any current and happening event all around the globe. In the e-commerce platforms, users often play a vital role in submitting feedbacks which help companies and franchises with invaluable data that can be used to manipulate their products as per the market scenario accordingly.
At the same time, it is quite obvious that an opinion or a review posted through any social networking site may not always directly state the exact orientation of the users but it may sarcastically express the true feelings. Sarcasm is a kind of sentiment which acts as an interfering factor in any text that can flip the polarity. If in a tweet or post, a word expresses a positive sentiment in a negative context, the tweet or the post should be classified as a sarcastic one. Unlike a simple negation, sarcastic tweets contain positive words. It is also true that an exaggerated variation of a positive word may convey a negative opinion or vice versa. So there is a requirement of analyzing this large volume of text in the form of reviews, tweets or feedback messages to gauge their precise orientation. To achieve this, each text may have to pass through a set of relevant algorithms.
However, finding and verifying the authenticity of opinions or reviews is not an easy task. Also, it is hard to comprehensively analyze all the reviews and recognize sarcasm in them. This is specifically true in case of product reviews. Any error in analysis may end up misleading them. The usage of feedback information is necessary in data-driven sectors in correspondence to Artificial Intelligence, which leads to several security and privacy concerns. Block-chain Techniques used for implementing several security methods ensures the provision of the rightful interpretations of information to customers from companies and vice versa (Cuong, Kumar & Solanki, 2020).
The term ‘sarcasm’ is generally used to mock, often with satirical or ironic remarks, with a purpose to amuse and hurt someone, or some section of society, simultaneously. This can also reflect a state of ambivalence and can be expressed variably through speech and text. Within straight talks, sarcasm is deluded by facial expressions and body gestures while within audio or telephonic conversations, sarcasm can be inferred by alterations in tone of voice. Thus sarcasm is one of the challenging areas faced in the field of sentiment analysis especially in text. It is the most difficult to identify compared to other methods, mainly due to absence of the verbal tone, thereby causing a frequent misinterpretation in conversations mostly done via writing i.e. Letters, emails, social media posts etc. (Literary Devices, Definition and Examples of Literary Terms, n.d.).