International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 11 Issue: 05 | May 2024
www.irjet.net
p-ISSN: 2395-0072
Echoes of Sentiment: Unveiling Amazon Alexa Product Reviews with Decision Trees, Random Forests, and XGBoost Arkajit Banerjee Department of Statistics , Acharya Prafulla Chandra College , Kolkata , West Bengal , India ---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - In this study, we conducted an in-depth
organizations and individuals to understand public opinion, consumer sentiment, and market trends. By leveraging NLP techniques, sentiment analysis algorithms analyze text at scale, uncovering valuable insights that can inform decisionmaking processes across a myriad of domains. From market research and brand sentiment analysis to social media monitoring and customer feedback analysis, sentiment analysis finds applications in diverse fields, including marketing, finance, customer service, politics, and healthcare. Understanding the nuances of sentiment analysis and harnessing the power of NLP techniques are essential for unlocking actionable insights from textual data and gaining a competitive edge in today's data-driven landscape.
investigation into sentiment analysis of Amazon Alexa product reviews using a publicly available dataset sourced from Kaggle. Our research journey commenced with a meticulous Exploratory Data Analysis (EDA), unearthing nuanced insights and patterns inherent within the dataset. Leveraging visualization techniques, we illuminated key trends, sentiments, and correlations within the Amazon Alexa reviews landscape. A pivotal aspect of our analysis involved the generation of word clouds, distinguishing between positive and negative sentiments. Through these visual representations, we meticulously showcased prevalent words that drove specific sentiments, enhancing the interpretability of our findings. Moving forward, we undertook rigorous data pre-processing steps, encompassing stemming using the Porter Stemmer algorithm and uniform conversion of all characters to lowercase. Subsequently, we ventured into the realm of sentiment analysis model development, employing three of the most renowned machine learning algorithms: Decision Trees, Random Forests, and XGBoost. Finally, we arrived at a pivotal stage of our study, wherein we evaluated the performance of our sentiment analysis models. Utilizing confusion matrices, we conducted a comprehensive comparative analysis, scrutinizing the accuracy and efficacy of the Decision Trees, Random Forests, and XGBoost models. Our findings shed light on the strengths and limitations of each model, providing invaluable insights for future sentiment analysis endeavors. In summary, our research not only contributes to the burgeoning field of sentiment analysis but also offers actionable insights for businesses and stakeholders seeking to comprehend and leverage consumer sentiments within the Amazon Alexa product ecosystem.
This paper aims to shed light on the imperative role of NLP techniques in sentiment analysis and its far-reaching impact on customer reviews of Amazon Alexa Reviews . Sentiment analysis plays a crucial role in understanding user interactions and feedback within voice-activated virtual assistants like Amazon Alexa. As one of the most popular virtual assistants on the market, Amazon Alexa interacts with millions of users daily, processing voice commands and providing responses to queries across a diverse range of tasks, including setting reminders, playing music, controlling smart home devices, and providing weather updates. Sentiment analysis of user interactions with Amazon Alexa can provide valuable insights into user satisfaction, preferences, and areas for improvement, enabling Amazon and third-party developers to enhance the user experience and tailor responses to better meet user needs. By integrating robust machine learning techniques like Decision Trees, Random Forests, and XGBoost into the analysis of user sentiment interactions with Amazon Alexa, stakeholders can gain deeper insights into user sentiment and behavior in the rapidly evolving landscape of voice-activated technology.
Key Words: Natural Language Processing , Bag of Words , Porter Stemmer , Decision Tree , XGBoost , Random Forest Classifier .
The key take-aways from the paper will be the rigorous exploratory data analysis (EDA) as well as the modelling approach . The rest of the paper’s structure is delineated as follows: Section 2 delves into the pertinent background and literature survey. In Section 3, the experimental methodology is expounded, encompassing the elucidation of theories, methods, approaches employed. Section 4 elucidates the results and graphs derived from the experimental procedures. Section 5 provides insights into the Conclusion and Future Plans. The last section is about references .
1.INTRODUCTION In today's digital age, where vast volumes of textual data are generated daily across various online platforms, the need to extract valuable insights from this data has become paramount. In the ever-evolving landscape of natural language processing (NLP), sentiment analysis emerges as a critical facet, unraveling the intricate layers of human expression embedded in textual data. . Sentiment analysis, a subfield of NLP, involves the automated identification and classification of sentiments expressed in text, enabling
© 2024, IRJET
|
Impact Factor value: 8.226
|
ISO 9001:2008 Certified Journal
|
Page 817