/parsers

Primary LanguageJupyter Notebook

BUILD FOR GENZ

Typing SVG

Trend-Centric Recommendation System

Employ AI-driven recommendation systems that analyze current fashion trends, customer preferences, and purchase history to deliver tailored product suggestions, enhancing customer engagement and conversion rates in the fast fashion segment.

Solution

AI Assistant Recommendation System by WebScrapping Developing an AI Assistant which generates Trend-Centric Recommendations by using a Recommendation model developed by taking data from Myntra's website by Webscrapping using Selenium and generating solutions by checking the ratings,prices and needs of the user at the current time

Steps

  1. Web scrapping using Selenium by using Beautiful Soup. We can also scrape data by first trying to parse with lxml if it's not present try looking for something in the script tag, sometimes data is loaded as JSON in script if it's not there either try something else like Selenium/Playwright.
  2. Building a Trend-Centric Recommendation System
  3. AI Assistant to deal with the queries regarding fashion industry and current fashion trends.

Tech Stack

javascript python react redux react-router material-ui html css sass bootstrap figma canva flask

Setup of local environment

  1. Fork this repo
  2. Clone the repo: git clone https://github.com/VanadiumV/parsers.git.
  3. Data Preparation: All the data used for training and testing the models (web scraping and recommendation system).
  4. Run Web Scraping Notebook: Open WebscrappingusingSelenium.ipynb in Jupyter Notebook.
  5. Navigate to cloned repo.
  6. Install all the necessary packages and libraries.
  7. Run command python app.py to run the chatbot.
  8. Chatbot will run on bot : http://127.0.0.1:5000/
  9. To check out the predictions.
  10. Setup for recommendation analysis model
  11. Run the jupyter notebook recomendation_rating_clothes.ipynb

Future Plans

  1. Further improving the bot by providing the real-time data of Myntra, so as to further help user to get access of latest trends on Myntra along with the trends in the world.
    • This will not only help users but also helps sellers to further customize their products according to user to increase the sell.
  2. Further improving the accuracy of the recommendation system by dividing the into multiple parts:
    • Based on temporary trends influenced by festivals, influenced by actors, movies, etc.
    • Based on users purchase history.
    • Trending on Myntra.
    • New on Myntra.

Team Members

Name LinkedIn
Valleri Jayal Link
Sarita Link
Nidhi Mishra Link

Demo Video

Myntra.mp4
Screenshot 2024-07-15 at 12 55 45 PM