MARS: A Multi-Modal Wartime Distress Signal Detection System Using NLP, Machine Learning, and Automatic Speech Recognition
Keywords:
Distress signal detection, Machine learning, Natural language processing, Support vector machine (SVM), TF-IDFAbstract
In today’s digital era, a large amount of textual communication takes place through social media platforms, messaging applications, and online support systems. Identifying distress signals from such textual data is crucial for enabling timely intervention and emergency response. Manual monitoring of distress messages is inefficient, time-consuming, and prone to human error. Therefore, there is a need for an automated system capable of detecting distress signals accurately and quickly. This project presents a Distress Signal Detection System using natural language processing (NLP) and machine learning (ML) techniques. The system processes user-entered text and performs preprocessing steps such as tokenization, stop-word removal, and normalization. Relevant features are extracted using TF-IDF vectorization, and classification algorithms such as Naïve Bayes, Support Vector Machine (SVM), and logistic regression are used to determine whether the input text indicates distress or not. If distress is detected, the system generates an alert notification through an email-based alert mechanism. The proposed system aims to improve response time, enhance monitoring efficiency, and provide a scalable solution for detecting emergency-related textual communication. This system can be applied in areas such as social media monitoring, emergency helplines, mental health support platforms, and public safety applications.
References
J. Walterskirchen, “Text as data for crisis-early warning: A comparative assessment of NLP methods for conflict prediction,” Conflict Management and Peace Science, Mar. 2026.
M. M. Islam, I. H. Chowdhury, and T. Pal, “Crisis communication on social media: A natural language processing and machine learning analysis of organizational responses and stakeholder engagement,” Saudi Journal of Engineering and Technology, vol. 10, no. 12, pp. 636–647, Dec. 2025.
J. Katalinić and I. Dunđer, “Neural network-based sentiment analysis and anomaly detection in crisis-related tweets,” Electronics, vol. 14, no. 11, pp. 2273–2273, Jun. 2025.
S. Das, K. S. Gogulamatham, K. Krishnaveni, M. K. Sharma, K. Namratha, and G. Kundanika, “Real-time crisis response and resource allocation using natural language processing,” International Journal for Research in Applied Science and Engineering Technology, vol. 13, no. 4, pp. 2611–2617, Apr. 2025.
P. Kanungo, “Natural language processing for social media sentiment analysis in crisis management,” World Journal of Advanced Research and Reviews, vol. 24, no. 2, pp. 2857–2864, Nov. 2024.
J. Liu, T. Singhal, L. T.M. Blessing, K. L. Wood, and K. H. Lim, “CrisisBERT: A robust transformer for crisis classification and contextual crisis embedding,” arXiv.org, 2020.
R. Lamsal, M. R. Read, and S. Karunasekera, “CrisisTransformers: Pre-trained language models and sentence encoders for crisis-related social media texts,” Knowledge-Based Systems, vol. 296, Jul. 2024.
A. Swaminatha, I. López, R. A. Garcia Mar, T. Heist, T. McClintock, K. Caoili, and M. Grace “Natural language processing system for rapid detection and intervention of mental health crisis chat messages,” npj Digital Medicine, vol. 6, no. 1, Nov. 2023.
N. Makkena, A. R. Islam, C. Varol and M. K. An, “Urgency detection in social media texts using natural language processing,” 2024 IEEE 18th International Conference on Semantic Computing (ICSC), Laguna Hills, CA, USA, 2024, pp. 156–163.
H. Senaratne, M. Mühlbauer, S. Götzer, T. Riedlinger, and H. Taubenböck, “Detecting crisis events from unstructured text data using signal words as crisis determinants,” International Journal of Digital Earth, vol. 16, no. 2, pp. 4601–4620, 2023.
