Everyone has different, perception, belief and feelings about product, services, destination, person etc. and is termed as Sentiments. SA is a Natural Language Processing (NLP) technique used to classify positive, negative, or neutral tones of a piece of text made on digital media. Further, the SA identifies specific emotions such as happiness, sadness, anger, fear and so on in regard to the person who made the comment. The importance of SA is emerging due to its capability to measure the quality perception, customer satisfaction and services.
As the world moving towards digital era, the way of expressing people’s feelings has been upgraded from papers to digital platforms. Thus, to analyze them, many Multinational Corporations (MNCs) use SA for Marketing, Product Development, Social Media Monitoring, Political Analysis, etc. For instance, Amazon being the giant ecommerce platform, employs SA to analyze customer reviews and feedback, identifying areas for product and service improvement and enhancements. They utilize it to assess the effectiveness of its marketing campaigns. Similarly, X (Twitter) a Social Media, practices it to monitor public sentiment regarding current events and potential trends, as well as to enhance the targeting of its advertising. Google a search engine where world’s information is organized, uses SA to optimize the quality of its search results and to track public opinion on recent news coverage. For 2020 US presidential election, Google used SA to see how people were reacting.
After COVID-19, Nepalese customers are shifting towards online platforms and making habit of giving feedbacks about it. But there aren’t any such tools to analyze data in Nepalese context. SA being a new topic in Nepalese market businesses can use it to analyze the customer reviews and feedbacks, ratings to identity the area of improvement ultimately satisfying the customers. As the technology keeps evolving and becomes more accessible, companies anticipate witnessing even more innovative and impactful uses of sentiment analysis in the Nepalese market.
Avant Garde Exploring the Trend in SA
SA being new context for Nepal, we are trying to step in towards analysis of digital data. On the process of study, we performed a small analysis on people’s perception towards Swayambhunath (Temple/Place) given on TripAdvisor (trekking agent’s portfolio). For this we scraped data from TripAdvisor, cleaned it and then classified it into positive, negative and neutral using RoBerta Model.
The results showed that 84% of the reviews were positive, 5% were negative, and 11% were neutral. An outliers were also observed in the distribution on the plot perception vis a vis rating. The study also identified the most frequently used words by creating a word cloud. These words included unigram words such as "Buddhist" & "beautiful", bigram words such as "highly recommended" & "beware monkeys", and trigram words such as "great views city" & "heritage site Kathmandu."
In conclusion, Sentiment Analysis (SA) holds substantial potential for improving customer satisfaction and business services. There is a vision for SA to emerge as a novel avenue for research organizations. As technology progresses, we anticipate the development of even more innovative and influential applications in the future