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{198769,
        author = {Dr. T. Ranjith Kumar and Pala Prathima},
        title = {Comparative Study on Rule-Based and Statistical-Based Information Extraction from Flipkart Customer Reviews},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {11},
        pages = {11462-11467},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=198769},
        abstract = {Information Extraction (IE) from unstructured text plays a significant role in transforming large volumes of raw customer feedback into meaningful, actionable insights. Since customer experience is becoming more important in determining business strategy, the ability to systematically extract sentiments and key feedback components from reviews has become a crucial task. This paper presents a comprehensive comparative study of two major techniques for information extraction — rule-based methods and statistical (machine learning) models — specifically applied to Flipkart smartphone customer reviews. The rule-based method uses predefined patterns, regular expressions, and manually selected keyword lists to identify sentiments and extract important aspects. On the other hand, the statistical-based approach utilizes advanced natural language processing (NLP) models, including transformer-based sentiment classifiers and Named Entity Recognition (NER), to automatically learn and infer insights from the review text without explicit programming of rules. For both methods, we systematically build and apply models, using a generated dataset of Flipkart customer reviews. A detailed evaluation is conducted based on precision, recall, F1-score and qualitative analysis of extraction accuracy. Comparative results highlight the strengths and limitations of each method, showing that rule-based techniques offer higher precision for predictable patterns, while statistical models provide greater adaptability and contextual understanding, especially in handling nuanced or mixed sentiments. Finally, the paper proposes a hybrid solution that leverages the precision of rule-based extraction and the flexibility of statistical methods, providing a robust framework for practical large-scale customer review analysis.},
        keywords = {Information Extraction, Sentiment Analysis, Rule-Based Systems, Machine Learning, Named Entity Recognition, Flipkart Reviews.},
        month = {April},
        }