Semantic Search based Movie Recommendation system.
In contrast to traditional search methods, semantic search is a method of extracting relevant information from a large corpus of documents.
This repository is an attempt to illustrate the same.
- Backend: Flask, Jina
- Frontend: Html , CSS, Javascript , React.js
- Clone the Repository
https://github.com/Akshit17/MovieRecommender.git
- Create a virtual environment
virtualenv myenv
- For Linux
source ./myenv/bin/activate
- For Windows
./myenv/Scripts/activate
- Install all the required packages with pip
pip3 install -r requirements.txt
💢 Make sure that the dataset in ./data_finalize/movies_metadata.csv
is indexed by running jina_app.py
with line 55 Uncommented.
Dont Forget to Comment it again after indexing to avoid indexing everytime a query passed when integrating it with Flask app.💢
- Run
jina_app.py
(Again after indexing the data from previous step) to serve Flow as a service . {Note down the address at which it is served from CLI for eg. "0.0.0.0:34567"}
python3 jina_app.py
💢Make Sure in app.py
line 44 url
is same as where Flow is being served so that Flask app and Jina can communicate💢
- Run the Flask Application
app.py
in another terminal alongsidejina_app.py
to get the web applcation running at "localhost:port"
python3 app.py
🍙 🍟 You've made it to the bottom 🧁 🍕