About my ML project

Introduction

The quick info about what I had done in this project is to created a model that recommend movies that related to the selected movie. for that I created model using cosine similarity. and deployed the model by using streamlit library. you can check recommended similar movies list for avatar by above picture.

Contents

1.Data collection

2.Data preprocessing

3.Model creation

4.Model deployment

5.Result prediction

Data collection

This data set is downloaded from the kaggle here is the link for the dataset

https://www.kaggle.com/tmdb/tmdb-movie-metadata?select=tmdb_5000_movies.csv

https://www.kaggle.com/tmdb/tmdb-movie-metadata?select=tmdb_5000_credits.csv

Data preprocessing

The first most thing is to load the data sets in to my jupyter environment. so that i can able to perform all the necessary data preprocessing techniques that i can get rid of outliers ,nanvalues, un necessary features , converts feature into appropriate lists. also merged two data sets .also performed stemming for un necessary words from using nltk library and many other .. I cleaned the data.. You can check this step from my git hub link https://github.com/prashanthdeva27/Movie-Recommendation-System

Model creation

In this step first i convert all my text data into vectors for that i need to import count vectorizer from sklearnlibrary. then i imported the cosine similarity from sklearn lib and then i fit the model and build the model. predicted recommendations. you can checkout the all the code necessary code from git hub link

Cosine similarity

Cosine similarity is a measure of similarity between two nonzero vectors of an inner product space that measures the cosine of the angle between them. Cosine of 00 is 1 and it is less than 1 for any other angle:

Here, Ai and Bi are components of vector A and B respectively:

Model deployment

I deployed this model by using streamlit library . I loaded the all necessary pickle files and referred some documentation in the streamlit website for creating a web app takes code from the streamlit website. link for streamlit documentation https://docs.streamlit.io/en/stable/ you check can this app code from my git hub https://github.com/prashanthdeva27/Movie-Recommendation-System/blob/main/app.py

Results

results can be predicted running the app by using code in command prompt : streamlit run app.py . you can check the necessary code from git hub link given https://github.com/prashanthdeva27 here is some images of app

Conclusion

this is my first step towards creating projects using ml algorithms by my own self study . I successfully built and deployed the model like an end to end project. iam facing some issues with deployement in heroku. i try to fix this as soon as possible you can contact if any necessary corrections and suggestions regarding this project. email:prashanthdeva27@gmail.com

Thank you..........