Skip to content

Latest commit

 

History

History
61 lines (51 loc) · 2.57 KB

README.md

File metadata and controls

61 lines (51 loc) · 2.57 KB
Raw

izooslorz

Tugas Besar Algeo II is a simple seach engine. Using cosine similarity, this project can find you similar document. It's also provided to add new documents from local files, link (web scrapping), and manual input.

Table of contents

General info

A Search Engine using TF and IDF with Cosine Similarity.

Requirements

To run this on your pc, you need:

  • Python 3.7.X (Recommended 3.7.9)
  • NPM and Node.js

Technologies

  • Flask
  • Textract - Extract Content from All Document Extension
  • NLTK - Stemming & Stopword for English
  • Sastrawi - Stemming & Stopword for Bahasa Indonesia
  • React
  • @material-ui

Setup

First, you need to install pip :

  1. Download "get-pip.py" file on pypa.io
  2. Run cmd as Administrator
  3. Type in the following command : $ python get-pip.py

Second, you need to install the dependencies. All the required dependencies can be found in the folder server by the name "requirements.txt" using $ pip -r requirements.txt

Third, you need NPM and node to start the development server for the client side/front end.

Fourth, you need to install the dependencies to start the client server. The dependencies are as following :

  1. Go to client directories, @material-ui/core, which can be installed by going to the directory of this project on cmd ran as administrator and by typing in $ npm install @material-ui/core
  2. @material-ui/icons, which can be installed by going to the directory of this project on cmd ran as administrator and by typing in $ npm install @material-ui/icons
  3. axios, which can be installed by going to the directory of this project on cmd ran as administrator and by typing in $ npm install axios
  4. react-router-dom, which can be installed by going to the directory of this project on cmd ran as administrator and by typing in $ npm install --save react-router-dom
  5. Run $ npm run start
  6. Go to server directories and run backend server using $ py app.py

Features

List of features ready and TODOs for future development

  • Ready: Web scraping by posting an url to the server
  • Ready: Search documents (pptx, docx, doc, pdf, txt, html, json) by getting a keyword to the server
  • Ready: Bahasa Indonesia and English supported
  • TODO: Change CSV database to SQL database such as Postgresql or mysql
  • TODO: Improve code efficiency
  • TODO: Improve frontend UI/UX

Inspiration

Project inspired by google, based on cosine similarity theory.