STRUCTURED
DATA
EXTRACTION
FROM
DOCUMENTS

Abstract

The International Federation of Red Cross and Red Crescent Societies assigns funding to National Socieities in the event that a disaster has occurred. This often has to be decided based on previous disasters that have occurred in the region, how frequently they have occurred as well as how much funding has been requested in the past in addition to other information.


Unfortunately, the structured historical data on past disaster events is limited and therefore meaningful historic trends are difficult to form. Presently, the IFRC has a platform through which National Societies can make Appeals. This platform collects relevant information from what has been submitted. However, this does not account for all the Appeals that were made manually in the form of documents from 2006-2020.


Our tool allows users to upload documents, performs the necessary enrichments and extractions, displays these to the user and stores the extracted information in a database that the IFRC can monitor and manage.


Main Features

Named-Entity Recognition

Extraction Tool

Question Answer Model

String Matching

PostgreSQL Database

Project Video


Team Members

zayn

Zayn Baig

Team Manager, Client Liaison, Full-stack Developer

Tosin

Tosin Fadahunsi

DevOps, Back-end developer, Back-end Integration

Tim

Stone Chen

UI Designer, Front-End Developer

Tim

Hannah James

Team Manager, Client Liaison, Back-end developer


Gantt Chart

Gantt Chart Part 1