Amazon consumer’s product preferences and sentiments – Analysis of product reviews using machine learning
Program: Data Science Master's Degree
Location: San Jose, California (remote)
Student: Raju Nadimpalli
The project will evaluate the significance of online reviews concerning Amazon products, background, project objectives, project goals, research methodology being considered, and the anticipated challenges. The analysis is primarily applied to cover consumer product preferences and capture sentiment from the text reviews. The study is limited to electronic products only.
One of the primary objectives is to predict the sentiment from unstructured data and validate the results with actual information. Periodically validate the model against the current data and optimize as and when needed. The second objective is to identify the consumers’ preferred and least preferred products, which could help Amazon improve its product offerings.
Source data collected from Kaggle will be subjected to exploratory data analysis. Unnecessary data will be cleaned up from the data set, including NAN, punctuation, syntax, stop words, etc. Each review text will be separated into words to identify the most used in the reviews. Created a word cloud to highlight the most frequently used words.