Project: Improving Natural Disaster and Water Resource Management using NLP
Overview
As a Machine Learning Engineer at Omdena, I contributed to a two-month challenge aimed at enhancing Water Resource Management (WRM) by leveraging space data. The project addressed the impact of natural calamities, such as droughts and floods, on water resources.
Problem Statement
The goal was to ensure sustainable water availability for drinking, sanitation, food production, energy generation, and industry. The challenge involved developing innovative solutions for WRM amidst population growth, urbanization, changing dietary habits, and climate uncertainties.
Project Outcomes
- Built a multifaceted NLP tool enabling users to query critical information from the internet.
- Users could define the topic, area, and time for data gathering, supporting queries like flood evaluations, drought indications, landcover mapping, and urban climate insights.
- Developed an efficient data storage and evaluation system, providing statistical insights on the gathered information.
Achievements
- Created an NLP tool for improving disaster response and water resource management.
- Deployed the Streamlit application on Google Cloud Platform Docker.
- Conducted web scraping using the GooglePyNews package for flood-related events.
- Designed the deployment architecture aligning with business requirements.
Recommendations
- **Inuri Illeperuma (Product Manager)
- **Rosana de Oliveira Gomes (Senior Data Scientist)
Project Details
For more information about the internship project, you can visit Omdena’s AI Water Management Project Page.
This experience enhanced my machine learning and data engineering skills, emphasizing the practical application of technology for societal and environmental impact.