🖌️ Interior Design Recommender

This project holds the source for the recommender system created for livspace hackathon

❓ What is recommender system?

• A system that profiles customers interior design requirements.
• The system matches the customer profile with designs by classifying and tagging designs and recommends desings.

⚙️ What is this built with?

1. Python3
2. Django
3. MySql

🚀 How to get started?

• Clone this repo

https://github.com/abhisekpadhi/livspace.git

• Setup virtualenv

cd insuretech_bot/
virtualenv --python=python3 venv

• Activate virtualenv

source venv/bin/activate

• Install dependencies

pip install -r requirements

• Setup database

mysql -u <username> -p<password>
mysql> CREATE DATABASE IF NOT EXISTS `livspace`;
mysql> exit

• Run migrations

python manage.py makemigrations
python manage.py migrate

• Dump dummy data (optional)

mysql -u <username> -p<password> livspace < livspace_data.sql

• Start application

python manage.py runserver localhost:8000

😃 API's

Questions API

• User requirement profiling with a questionnaire.

curl -X GET \
  http://localhost:8000/api/questions/ \
  -H 'Content-Type: application/json'

response:

{
    "status": "success",
    "message": {
        "questions": [
            ...
            {
                "question": [
                    {
                        "id": 1,
                        "text": "What is your age bracket from following ?",
                        "type_for_customer": "single-select",
                        "name": "age",
                        "type_for_designer": "multiple-select"
                    }
                ],
                "choice": [
                    {
                        "id": 1,
                        "question": 1,
                        "text": "30-50",
                        "is_inactive": false
                    },
                    {
                        "id": 2,
                        "question": 1,
                        "text": "50-70",
                        "is_inactive": false
                    },
                    {
                        "id": 3,
                        "question": 1,
                        "text": "70+",
                        "is_inactive": false
                    }
                ]
            }
            ...
        ]
    }
}

• Receive answers

curl -X POST \
  http://127.0.0.1:8000/api/questions/ \
  -H 'Content-Type: application/json' \
  -d '{
	"user_id": 1,
	"tags": [{"question": 1, "choice": 1}]
}'

response

{
    "status": "success",
    "status_message": "User created."
}

• Updata a user

curl -X PUT \
  http://localhost:8000/api/v1/users/ \
  -H 'Content-Type: application/json' \
  -d '{
	"email": "test@google.com",
	"phone_number": "+91943983125"
	}'

response

{
    "status": "success",
    "message": {
        "designs": [
            {
                "id": 1,
                "image_url": "http://127.0.0.1:8000/static/images/Vittoria-Orlandi-6.jpg"
            },
            {
                "id": 3,
                "image_url": "http://127.0.0.1:8000/static/images/Vittoria-Orlandi-6_JoAAJqx.jpg"
            },
            {
                "id": 4,
                "image_url": "http://127.0.0.1:8000/static/images/Vittoria-Orlandi-6_lC4nMY5.jpg"
            },
            {
                "id": 5,
                "image_url": "http://127.0.0.1:8000/static/images/3D-Interior-Design-Service-Contractorbhai-Modern-Design-4-400x265.jpg"
            },
            {
                "id": 6,
                "image_url": "http://127.0.0.1:8000/static/images/T.jpg"
            },
            {
                "id": 7,
                "image_url": "http://127.0.0.1:8000/static/images/M.jpg"
            },
            {
                "id": 8,
                "image_url": "http://127.0.0.1:8000/static/images/royal-bedroom.jpg"
            }
        ]
    }
}

Designer API

• Build desing tags mapping with input from desinger.

curl -X POST \
  http://10.2.41.106:8000/api/designers/ \
  -H 'Content-Type: application/x-www-form-urlencoded' \
  -H 'content-type: multipart/form-data; boundary=----WebKitFormBoundary7MA4YWxkTrZu0gW' \
  -F image_file=@/Users/abhisek.padhi/Desktop/livspace/royal-bedroom.jpg \
  -F user_id=1 \
  -F 'tags[]={"question": 1, "choice": 1}'

response

{
    "status": "success",
    "message": "Updated design tag mapping."
}

👌 ER Diagram and Flowchart