The paper analyzes sentiment of online food reviews using big data analytics to overcome limitations of traditional processing methods. It explores sentiment analysis techniques using Apache Spark machine learning libraries on a large Amazon Fine Food Reviews dataset. The author finds that a linear support vector classification model achieves over 80% accuracy and outperforms naive Bayes and logistic regression models. While support vector classification reduces processing time, future work could explore different n-gram feature sets to improve classifier performance.
The paper analyzes sentiment of online food reviews using big data analytics to overcome limitations of traditional processing methods. It explores sentiment analysis techniques using Apache Spark machine learning libraries on a large Amazon Fine Food Reviews dataset. The author finds that a linear support vector classification model achieves over 80% accuracy and outperforms naive Bayes and logistic regression models. While support vector classification reduces processing time, future work could explore different n-gram feature sets to improve classifier performance.
The paper analyzes sentiment of online food reviews using big data analytics to overcome limitations of traditional processing methods. It explores sentiment analysis techniques using Apache Spark machine learning libraries on a large Amazon Fine Food Reviews dataset. The author finds that a linear support vector classification model achieves over 80% accuracy and outperforms naive Bayes and logistic regression models. While support vector classification reduces processing time, future work could explore different n-gram feature sets to improve classifier performance.
paper is to analyze “Amazon Fine Food Reviews” which are growing exponentially that may not be processed by the system so to overcome this we apply big data concepts in it
To overcome those issues we
Important claim explore different methods and techniques of sentiment analysis using machine learning libraries of “Apache Spark” to process huge datasets of bigdata analytics
The main belief that the
Simplifying author made in this paper is assumptions made by the that to get the accuracy of 80% output they use three author mechanisms such as Linear SVC, Logistic Regression, and Naïve Bayes among he realize that Linear SVC performs more efficiently than NB and logistic regression. Work done by the author First he collects all the datasets from the “Amazon” then pre-process the data and then tries to analyze the various important aspects that use to describe the sentiment analysis such as three aspects are implemented 1. positive 2. negative 3. Neutral by splitting into these ways the analysis made
Evaluate their work with When compared with others
others the author finds that the linear SVC could give more accurate value when compared with others such as NB and Logestic Regression .
Limitations of their SVC reduces the processing
research time compared with other ML algorithms
Future research To improve the performance
of the classifier different feature sets will be considered such as bi-gram, tri-gram, and four-gram.