Volume 2 - Issue 2, March - April 2026

As digital communications continue to grow rapidly from one decade ago to present. Emotional detection from text becomes one of a few key areas of interest. Recognizing the manner in which people express emotions in their electronic writings allows for the development of intelligent systems, which will allow for the most effective response to electronic inquiries and/or to create an enhanced experience for the end-user of the system. Accurately identifying emotions from written messages will provide opportunities to create new applications in the fields of sentiment analysis, mental health monitoring, and CRM (Customer Relationship Management). A great deal of work has been done in developing intelligent approaches (i.e., neural networks) for emotional detection on text. We present a relatively simple method for the construction of a neural network, which enables accurate classification of text into five distinct emotional states: happiness, sadness, anger (including rage), fear, and neutral/undecided. Additionally, we describe ways to improve the proposed model's capacity to accurately classify emotional states by leveraging NLP and methods of preprocessing and featureextraction. Finally, we demonstrate that basic neural networks when used in conjunction with effective preprocessing techniques yield competitive results, thus demonstrating the potential to deploy basic neural networks for such tasks with lower cost and a high level of accuracy.

Patel Priyanshi,"Emotion Detection Using Simple Neural Network and Text Data" International Journal of Advanced Multidisciplinary Research and Educational Development, V2(2): Page(449-451) Mar-Apr 2026. ISSN: 3107-6513. www.ijamred.com. Published by Scientific and Academic Research Publishing.