Skip to content

abdullahwaqar/docsearx

Repository files navigation

docsearcx | A minimal document search engine.

alt text

docsearcx is a simple search engine that retrieves information from pdfs based on term frequency-inverse Document frequency and cosine similarity to retrieve relevant documents.

Limitation

For the sake of POC this application relies on in memory storage.


Setup

Installing Pipenv

If pipenv is already installed skip this step.

pip install pipenv

Installing Dependencies

pipenv install

& Activate the virtual environment shell by

pipenv shell

Running the Flask app

python app.py

Running Client

cd client/

npm install

npm run serve

Term Frequency-inverse Document Frequency

TF-IDF is a numerical statistics which reflects how important a word is to a document. The tf-idf value increases proportionally to the number of times a word appears in the document, but is offset by the frequency of the word in the corpus, which helps to control for the fact that some words are generally more common than others.

About

A simple search engine that ranks pdfs based on search keyword & TF-IDF weights and cosine similarity.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published