Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{203903,
        author = {Piyush Bhardwaj and Navin Prasad Jha},
        title = {Customer Sentiment Analysis from Big Data},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {13},
        number = {1},
        pages = {692-696},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=203903},
        abstract = {In the fast-digitalize age, consumers are increasingly leaving their words and experience on every corner of the Internet e-commerce platforms, Instagram feeds, Reddit channels, and so on. Sometimes it is no laughing matter to wade through all that disorganized talk but it is crucial to determine what people really want and make more prudent business decisions. This project consists of a machine learning sentiment analysis pipeline that identifies a positive, negative or neutral review. We begin by feeding the text into an NLP stack to clean, tokenize and normalize, then extract the high- value bits with TF-IDF, and lastly toss it to a logistic regression classifier because, it is simple, easy to read and also it works on data sets of medium size. The data set I was working with is simply huge heap of customer reviews. I read them all, removed noise, tokenized, and ensured that the words were uniform enough so that the model would be able to learn correctly. In order to evaluate the performance of the model, we struck it with the standard suspects, namely, its accuracy, precision, recall and the F1 -score. The figures emerged solidly at about 90 per cent accuracy, that is, the projections are most reliable and repeatable. In addition, I have packaged all that with a Streamlet web application that allows viewing the sentiment scores in real-time, which comes in convenient when using it in the real business. Overall, the work demonstrates that traditional ML methods and intelligent feature extraction are not quite powerless in terms of the precision and resource- saving. The platform I have constructed is readily scalable, and the subsequent measures may include connecting it with a big- data system or adding a layer of deep learning to push the performance to the extreme.},
        keywords = {Customer Sentiment Analysis, Natural Language Processing (NLP), Machine Learning, TF-IDF, Logistic Regression, Text Classification, Big Data Analytics, Opinion Mining, Customer Reviews, Stream lit Application},
        month = {June},
        }