Exploring consumer opinions on vegetarian food by sentiment analysis method

Abstract

The study’s aim is to explore consumer opinions on vegetarian food in the electronic commerce area by the qualitative research method, in more detail, based on utilizing the lexicon-based approach to analyze sentiment. The sentiment of five aspects including price, package, shipment, brand, and quality is considered detecting from customers’ reviews. Besides, the rating tags are also categorized as an opinion (positive, negative, neutral) based on the star number. The dataset includes 17,892 customer reviews. The findings found that the customers’ sentiments are 20.8% positive, 0.7% negative, and 78.5% neutral for an average of five aspects. Product quality is the most concerning of the five aspects of customer comments. This aspect also makes up 8,672 positive comments. Besides, quality and package aspects are more than 5,000 positively attached tags and shipment has the most negative tags taking 26 rating tags. Furthermore, the average of the stars is 4.94 stars, which is close to reaching the peak of 5.0 stars. This study demonstrates that the view of customers for aspects about vegan products according to text comments, the quality aspect has a significant number of positive comments than other aspects. Hence, the business might understand the significant factors to deal with them.

Keywords

• lexicon-based approach
• product review
• sentiment analysis
• vegetarian food
• vietnamese market