Volume 30, Issue 1, 2021
DOI: 10.24205/03276716.2020.2065
Amazon Product Sentiment Analysis using Machine Learning Techniques
Abstract
Online stores like Amazon provide a website for consumers to express their opinions about different items. Since then, it has been established that buying online, 90% of consumers are testing different websites channels to determine the quality of their purchase. To evaluate the text data and then extract the sentiment element from that the field of sentiment analysis is frequently used. From user ratings, suggestions, recommendations and messages, online business websites produce a massive volume of textual data every day. One of the most emerging technological trends in web development is the emergence of social web sites. It helps to communicate the peoples and gather knowledge. It simplifies the user contribution via podcasts, blogs, folksonomies, tagging, and wikis using online Social Networks (OSN). Social Network Analysis (SNA) methods can cover extensive networks from thousands or millions of nodes and links of a graph. It characterizes the network structure in the form of nodes (specific actors, people, or things) that is a network used or within the network and the edges or links between nodes (relationships) that connected the network. In this research study, we presented a novel approach that uses sentimental aspects focused on the characteristics of the item. Amazon consumer reviews have been introduced and validated. We collected the dataset from the data world centre, where opinion rates are first detected in each analysis. The system performs pre-processing operations like stone-coating, tokenization, boxing, deletion of stop-words from the datasets to extract meaningful information like positivity or negativity. The primary study aims to analyze this data's aspect level is a huge benefit to marketers to grasp the preferences of consumers better and then develop their behaviour accordingly. Lastly, we also give some perception into their future text classification work.
Keywords
Machine learning, social network analysis (SNA), amazon, polarity, user reviews.