Workshop Recommendation from Customer Feedback Using Sentiment Analysis

Abstract

Abstract—OnRoad vehicle breakdown assistance finders help find the nearest mechanic shop that let them book services for their vehicle work. But in this process of choosing nearby mechanic shops, they’re unaware of the fact whether the shops they go actually provide best services or not and in order to know they have to conduct verbal conversations with people for reviews which is technically not a feasible option. Here comes the potential need for a strong recommendations system that generates feedback based on user reviews through sentiment analyses that let them choose the right workshop. The systematic identification of emotions such as positivity, negativity, or neutrality in textual data is involved in sentiment analysis, which is also known as opinion mining. Subjective data is extracted and analyzed using text analysis and NLP in this technique, which finds applications in a variety of fields such as marketing, customer service, and clinical medicine. The analysis of service-based feedback in the paper focuses on determining users' level of satisfaction or dissatisfaction with the mechanic service by analyzing their emotions. It is possible to quickly identify the emotional state of customers and the factors that contribute to it by evaluating their feedback. Hybrid approach which is a combination of lexicon approach and TextBlob approach using Naïve Bayes algorithm. TextBlob approach helps in identifying misspelled words and providing potential solutions is the process of correcting them in a text. In a document, a Spell Checker tool is usually utilized to identify misspelled words. The program Spell Checker scans the dictionary to identify words spelled incorrectly and proposes replacements.