Customer Review Analysis and Identifying Spam Reviews

Abstract

Gathering customer feedback is essential for any organization, especially in the airline service sector. Surveys are one of the most common ways to collect customer feedback and measure customer satisfaction. However, creating surveys and managing survey responses can be a challenging and time-consuming task. This is where twitter come’s into picture, where everybody can share their opinion, this allows airlines to understand how customers feel about their services, identify areas of improvement, and make necessary changes to improve their services. By using this, airlines can gain valuable insights into customer preferences that can help them create more personalized experiences for their customers. Additionally, it can help airlines stay ahead of the competition by understanding what customers want and providing them with better services than their competitors. To achieve a better overall rating from their consumers, some businesses employ comment spam to downgrade the rankings of their competitor firms based on the categories of their items. Thus, one of the tasks to boost sentimental analysis's authenticity is to analysed these spammers patterns and identify them as genuine or fake. Thus, in this project we focus on the reviews which is given by the people on twitter for an airline company and for every individual review. We basically classify them into spam or not spam thus we will use different algorithms like Support Vector Machine,Naïve Bayes, Random Forest and choose the optimal one after comparison of each algorithm