Name-Verification-Model

NLP project to detect typos in Arabic names.

Description 📜

The task is divided into multiple stages (Data generation - Core model - Interface - containerization). The core stage of the task pipeline is name verification model with confidence threshold of how much the name is real.

For example:

The name "باسم وحيد السيد" is a real name with high confidence.
The name "باسمم وحةد السد" “is a real name with low confidence.

Datasets Source Links 📁

The datasets source is the same source you provided to me in the task description PDF

https://drive.google.com/drive/folders/1rW6vtJ_niYLao1YrYD0D4psAaqt0tMyh

Instructions 📝

How to build the docker image

To build the Docker Image we should execute the following docker-compose command

docker-compose -f docker-compose.yaml up

After completion of building the last part in the output in the terminal should be as following

Jun 25 09:12:00 PM   * Running on all addresses (0.0.0.0)
Jun 25 09:12:00 PM   * Running on http://127.0.0.1:80
Jun 25 09:12:00 PM   * Running on http://10.223.28.167:80
Jun 25 09:12:00 PM  Press CTRL+C to quit

So, in this case the app has been built successfully!

How to test the solution

There is two ways to test the solution

At this point the app was built and the connection to the server is active now, so we can head now to the server by opening the browser and type this URL in the search bar

http://127.0.0.1:80

Another way you can head to the server, is to open the Docker Desktop Application if it was installed on your local machine, you can open it then on the left press on Containers, then you will see our container expand it, then head to the PORT(S) column, and you will see this port 80:80, then press on it and it will get you directly to the server to test
There is endpoint to test the model directly, from the following link

https://name-verification-model.onrender.com/

How to verify results claimed

After you open the server, open the GET Tab that its name is /Inference, then press on try it out to let you enter some inputs, after you enter your full name to test, press on Execute, after that in the Response body you will see the results of the Prediction of the model and the Execution Time, as if the name was correct the result will be The Full Name is: Correct, and if the name was incorrect the result will be The Full Name is: Incorrect

so enter your test name and see the results of the model and see if the predection was right :)