erfan-hamidi/smart-text-analysis-dj

SAD project

smart-text-analysis-dj

System Analysis and Design Project - Buali Sina university

Project Report link

-> this project have to main AI parts :

1. sentiment analysis (Persian and english support) : this Phase is based on CNN model and used IMDB Dataset of 50K Movie Reviews .
2. text summarization (Persian and english support) : this Phase is based on Bart model that is a Seq-to-Seq Pre-training model for Natural Language Generation, Translation, and Comprehension . our dataset is cnn_dailymail .
3. image retreieval system : this Phase is based on VGG16 Deep model .
4. spam sms/email detection : using ensemble learning using : 1- MultinomialNB() 2-LinearSVC() 3-DecisionTreeClassifier() with 98.5% accuracy

downlaod model :

english sentiment analysis model link (imdb dataset)

persian sentiment analysis model link (digikala dataset)

english text summarization model link (cnn-dailymail dataset)

english spam detection model link

AI part technical report video

Charts

plots of persian sentiment analysis model

plots of english sentiment analysis model

image retrieval output result :

## Installation

    $ cd smart-text-analysis
    $ python -m venv venv
    $ source venv/Scripts/activate
    (venv) pip install -r requirements.txt
    (venv) python manage.py makemigrations
    (venv) python manage.py migrate
    (venv) python manage.py runserver