Data Science for Early Disease Outbreak Detection (Dengue)

Data Science for Early Disease Outbreak Detection (Dengue)

 Gurala Siva Teja Reddy
Department of Computer Science and Engineering,
Koneru Lakshmaiah
Education Foundation,
Vaddeswaram, Andhra Pradesh, India 2200031538@kluniversity.in

Sirigineedi Bharath Bhushan
Department of Computer Science and Engineering,
Koneru Lakshmaiah
Education Foundation,
Vaddeswaram, Andhra Pradesh, India 2200031507@kluniversity.in

Movva Tanmayi
Department of Computer Science and Engineering,
Koneru Lakshmaiah
Education Foundation,
Vaddeswaram, Andhra Pradesh, India 2200030187@kluniversity.in

Duggempudi Manikanta Reddy
Department of Computer Science and Engineering,
Koneru Lakshmaiah
Education Foundation,
Vaddeswaram, Andhra Pradesh, India 2000031311@kluniversity.in

Mrs. Thejo Lakshmi Gudipalli
Department of Computer Science and Engineering,
Koneru Lakshmaiah
Education Foundation,
Vaddeswaram, Andhra Pradesh, India tejolakshmi.gudipalli@kluniversity.in

Abstract— Dengue fever, a mosquito-borne viral disease, impacts millions annually, with an estimated 390 million infections worldwide, particularly in tropical regions like India, Southeast Asia, and Latin America [1]. Delayed detection of outbreaks exacerbates healthcare system strain, increases mortality rates, and hampers effective containment efforts, especially in dengue-endemic areas where monsoon-driven mosquito breeding fuels rapid spread [11]. This research proposes a comprehensive data-driven platform to detect and manage dengue outbreaks early, leveraging advanced big data analytics and machine learning to transform traditional surveillance, which often relies on slow, manual case reporting [2]. The platform integrates real-time data from diverse sources—hospital electronic health records (EHRs), wearable devices monitoring fever and heart rate, public health case reports, environmental factors like rainfall and temperature, and social media posts about symptoms—to create a holistic view of outbreak risks [7, 8, 10, 11]. Machine learning models, such as Isolation Forest and Long Short-Term Memory (LSTM) networks, identify anomalies like sudden spikes in fever cases, achieving 93% accuracy in detecting early outbreak signals [3, 7]. Geospatial visualization tools, built with ArcGIS, map outbreak zones and mosquito breeding sites, enabling targeted interventions like mosquito control, with a 35% reduction in response times compared to conventional methods [5, 14]. Real-time alerts notify health officials via dashboards, SMS, and email, ensuring rapid coordination [4, 11]. Predictive models forecast resource needs, such as hospital beds and insecticides, with 91% accuracy, optimizing containment strategies [9, 12]. The system adheres to privacy standards (HIPAA, GDPR) and ISO 27001 for secure data handling, addressing ethical concerns in public health [13]. Tested on simulated datasets from India (2020–2024), the platform demonstrates faster detection, improved resource allocation, and enhanced collaboration among healthcare providers and authorities, offering a scalable, ethical solution for dengue management in high-risk regions [6, 15].

Keywords— Dengue, Early Outbreak Detection, Data Science, Machine Learning, Geospatial Analytics, Big Data, Public Health