An experimental study of text preprocessing techniques on user reviews
Sonia Rani, Tarandeep Singh Walia · 2023
Text preprocessing plays a vital role in extracting important information from unstructured raw data and improving the quality of raw data. Preprocessing techniques transform the raw data into a more understandable format. The popularity of online shopping has been increasing the number of product reviews on e-commerce websites. However, it is very challenging to understand these informal reviews written by non-professionals, which contain so much noise, abbreviations, and spelling errors. So, text preprocessing is a crucial and essential step to removing the noise from reviews and understanding the users’ feedback about the products. This study’s primary purpose is to experiment with several preprocessing techniques, such as noise removal, contractions, spelling correction, stemming, and lemmatization on text data, and analyze the challenges of some preprocessing methods.