Named Entity Recognition Using NLP

Authors- Keerthana R, Assistant Professor Waseemuddin

Abstract-Recent advancements in Natural Language Processing (NLP) have paved the way for its widespread application across diverse fields such as business, law, and healthcare. An essential component of any NLP project is text preprocessing, a crucial step that involves modifying text data before feeding them into machine learning models. Typically, text preprocessing encompasses tasks like cleaning, filtering, removal, and replacement of certain texts to enhance model accuracy, robustness, reduce data size, or ensure privacy preservation. Named Entity Recognition (NER) stands as a key NLP tool, tasked with identifying Named Entities within text, including names, organizations, addresses, numbers, and dates. In this study, we propose a novel preprocessing approach leveraging NER to identify named entities and subsequently utilize them to enhance accuracy and safeguard privacy, instead of discarding them or allowing them to contribute noise to our data. Through a series of experiments conducted on various datasets, including some collected in-house, we evaluated our approach’s efficacy in text classification tasks. Our findings demonstrate that incorporating this approach not only boosts classifier accuracy and reduces dimensionality but also effectively preserves privacy. This underscores the significance of leveraging NER in text preprocessing to optimize NLP applications across different domains.

DOI: /10.61463/ijset.vol.12.issue2.142