SMS Checker / Backend

The backend of this project provides a simple REST service that can be used to detect spam messages. We have extended the base project rohan8594/SMS-Spam-Detection, which introduces several basic classification models, and wrap one of them in a microservice.

The following sections will explain you how to get started. The project requires a Python 3.12 environment to run (tested with 3.12.9). Use the requirements.txt file to restore the required dependencies in your environment.

Training the Model

To train the model, you have two options. Either you create a local environment...

$ python -m venv venv
$ source venv/bin/activate
$ pip install -r requirements.txt

... or you train in a Docker container (recommended):

$ docker run -it --rm -v ./:/root/sms/ python:3.12.9-slim bash
... (container startup)
$ cd /root/sms/
$ pip install -r requirements.txt

Once all dependencies have been installed, the data can be preprocessed and the model trained by creating the output folder and invoking three commands:

$ mkdir output
$ python src/read_data.py
Total number of messages:5574
...
$ python src/text_preprocessing.py
[nltk_data] Downloading package stopwords to /root/nltk_data...
[nltk_data]   Unzipping corpora/stopwords.zip.
...
$ python src/text_classification.py

The resulting model files will be placed as .joblib files in the output/ folder.

Serving Recommendations

To make the models accessible, you need to start the microservice by running the src/serve_model.py script from within the virtual environment that you created before, or in a fresh Docker container (recommended):

$ docker run -it --rm -p 8081:8081 -v ./:/root/sms/ python:3.12.9-slim bash
... (container startup)
$ cd /root/sms/
$ pip install -r requirements.txt
$ python src/serve_model.py

The server will start on port 8081. Once its startup has finished, you can either access localhost:8081/apidocs in your browser to interact with the service, or you send POST requests to request predictions, for example with curl:

$ curl -X POST "http://localhost:8081/predict" -H "Content-Type: application/json" -d '{"sms": "test ..."}'
{
  "classifier": "decision tree",
  "result": "ham",
  "sms": "test ..."
}