Skip to content

Latest commit

 

History

History
40 lines (25 loc) · 921 Bytes

README.md

File metadata and controls

40 lines (25 loc) · 921 Bytes

Search-Engine

Search Engine project for CSI4107 - Information Retrieval. Search engine built on top of UofO courses collection and the reuters text collection.

Quick Start

Install dependencies

Required packages are:

  • Python 3.7
  • Kivy
  • Nltk
  • Numpy
  • Pandas
  • BeautifulSoup

Note: Python 3.8 not supported as Kivy does not support that version yet

pip install -r requirements.txt

Preprocessing data and setting up indexes

Next setup dictionaries and indexes by running command below which should take around 3 minutes. Additionally you can add the --knn flag to run knn algorithm on reuters to predict topics. This will take around 8 minutes total. All preprocessed data and models are already included within the repo in collections and models.

python make_data.py
python make_data.py --knn #For knn on reuters

Start app

python search_engine_app.py