Abstracts Features Project Video Team Members Gantt Chart
Project Background Project Goals Gathering of Requirements Personas MoSCoW Requirements
Named Entity Recognition SpaCy Q&A Model Table Extraction String Matching Resources
UI Design System Architecture Tool Map Data Storage
Implementation Overview Retrieving and reading file(s) Answer Etraction Location Extraction Code Matching Display and save
Testing Strategy Unit Testing Integration Testing User Testing
Achievements QA Models spaCy Models String Matching Critical Evaluation Future Work
Deployment Manual Monthly Videos Development Blog

Design

UI Design

* note that the UI was optimised for Windows OS, amd thius should be run on Windows. It was not a requirement from our clients for it to work with other operating systems, as they had specified that they are Windows based.

Influences

When designing the app, we looked at many similar projects that involves same functionalities. And we drew inspiration from those projects.

There are numerous apps or websites that has the functionality to upload files either by drag and drop, or click and select. The above picture is from the cloud service Box which is one of the place where we looked at during the design process.

The above example is from css tricks, a website that does tutorials on css. We have decided to remove the upload button and make it even simpler, and users can figure out where to click easily.

We also made sure that the style of our app matches the website of IFRC.

Prototype

Main page

When the user clicks preview, the view page will be shown and the user can click right and left to view all the files uploaded to make sure they are correct.

The load screen will last until the extraction is completed.

The data extracted will be displayed and can be edited.

We have always kept “KISS” (Keep It Simple and Stupid) in mind when designing the UI, and therefore out UI only has the necessary functionalities and it is easy to use.

System Architecture

Tool Map

All users arrive on the home page. This is when they are prompted to select the documents they want to run. Once selected, they have the option to view these documents and then have them uploaded.
Once uploaded, the data retrieved from the documents are viewed individually. This is then followed by the option to edit the data and upload them to the postgres database.

Data Storage

The system contains a Postgres database consisting of 4 tables.
The admin 0 code is the Primary Key for the first table, The admin 1 code is the Primary Key for the second table and The admin 2 code is the Primary Key for the third table. The Disaster table uses the operation_number as the key as it is unique for all documents and references the other tables through the admin 2 code.