How I Built a Simple Movie Recommender Website (Without Hollywood Budget)

Ever opened Netflix, spent 40 minutes scrolling, and then closed it without watching anything?

Yeah. Me too.

That’s why I decided to build a simple movie recommendation web app—a clean, lightweight tool to help users discover movies based on their favorite genres.

---

Project Goals

• Recommend similar movies based on user input

• Keep it simple, minimal, and fast

• Use basic ML, not deep learning (yet)

• Learn Flask + deploy a real working web app

---

Tech Stack

• Python (core logic + ML)

• Flask (web framework)

• Scikit-Learn (vectorizing + similarity)

• HTML/CSS (UI)

• TMDB Dataset (from Kaggle)

• Hosted with Netlify (frontend) or Render (backend API)

---

The Building Blocks

1. Getting the Data

I started with a movie metadata dataset from Kaggle that included:

• Movie titles

• Genres

• Tags

• Overview descriptions

I cleaned and preprocessed it using pandasand selected a few key features to combine into a “bag of words.”

data['tags'] = data['overview'] + data['genres'] + data['keywords']

2. Vectorizing the Data

To compare movies, I used CountVectorizer from scikit-learn to convert text into a numeric form:

from sklearn.feature_extraction.text import CountVectorizer
from sklearn.metrics.pairwise import cosine_similarity

cv = CountVectorizer(max_features=5000, stop_words='english')
vectors = cv.fit_transform(data['tags']).toarray()

similarity = cosine_similarity(vectors)

3. Recommendation Logic

Simple cosine similarity:

def recommend(movie):
    index = data[data['title'] == movie].index[0]
    distances = similarity[index]
    movie_list = sorted(list(enumerate(distances)), reverse=True, key=lambda x:x[1])[1:6]

    for i in movie_list:
        print(data.iloc[i[0]].title)

Later, I wrapped this into a function used in the Flask app.

4. Building Frontend

Although I don’t barge too much behind the frontend, I tried to give my application a basic clean UI, with the knowledge I gained by harsh training( learnt in uni class).

<form action="/recommend" method="post">
    <input type="text" name="movie" placeholder="Enter a movie...">
    <button type="submit">Recommend</button>
</form>

Integrating Flask Backend

My Flask app had a basic structure:

from flask import Flask, request, render_template
app = Flask(__name__)

@app.route("/", methods=["GET"])
def home():
    return render_template("index.html")

@app.route("/recommend", methods=["POST"])
def recommend():
    movie = request.form['movie']
    result = recommend_movie(movie)
    return render_template("recommend.html", movies=result)

See, it's basic—just simple GET and POST requests, and wham! The application is ready.
(Although I did have to spend some time figuring out why it wasn’t serving the CSS stylesheet. Flask static routing can be tricky at 2 am.)

Lesson learned: always link your CSS via the proper /static/ folder and use url_for('static', filename='styles.css').

What I Learned

• Clean data = smart app

• Cosine similarity is simple but powerful

• Flask is great for quick MVPs

• UI design matters (even for tech demos).

Try It Yourself!

Code: GitHub Repository

Future

In the future, I would like to make some enhancements:

• Add content-based + collaborative filtering

• Add user login + saved watchlists

• Maybe a React frontend later.

Final Thoughts

It’s not IMDb, and it won’t win an Oscar — but it works.
This project taught me that sometimes, building something simple and functional is better than overengineering.

If you’re learning Python or Flask, this is a great starting point.
And if you ever feel stuck... just pick a movie and start coding

That’s enough larping for me, till next story. bye.. bye..