A new type of search engine.
Many seacrh engine like google,bing etc offer a better , robust way of searching any data from the internet.In this project we prupose a model that search or ranks pages based on the simantic simlarity between the search query and the data in the website(for instance all the headers and paragraps in the page)
Using the concept of web scraping and NLP combined to produce the final results.
Universal sentence encoder + cosine similarity
To Run
scrapy crawl getData -o data.json
a data.json file is created with the scraped items.
open the jupyter notebook
to run the remaining code