I used the TMDB Movie Metadata dataset from Kaggle, which comprises data on as much as 5000 films. The dataset consists of attributes like:
- Title
- Genres
- Overview
- Forged
- Crew
This wealthy dataset allowed me to leverage vectorization methods to research film descriptions and options.
To construct the system, I used the next:
- Python: For scripting and information processing.
- Pandas: To govern and preprocess the dataset.
- Streamlit: To create an interactive net interface.
- Vectorization: To compute similarities between film options utilizing methods like TF-IDF.
- IMDb API: To fetch film posters dynamically, enriching the suggestions visually.
After constructing the online app utilizing Streamlit, I deployed it on Hugging Face Spaces. Hugging Face Areas is a superb platform for internet hosting and sharing machine studying fashions and purposes.
You possibly can discover the app right here: Movie Recommender System
Like this Hugging Face area? Be happy to strive it out and share your suggestions!
Here’s a pattern video :
To make the suggestions extra partaking, I built-in the IMDb API from RapidAPI to fetch film posters dynamically. These visuals present customers with an enhanced shopping expertise.
Take a look at the IMDb API documentation here.
Creating this film recommender system was a rewarding expertise. It not solely deepened my understanding of machine studying methods but in addition allowed me to discover sensible deployment methods. Whether or not you’re a developer or a film fanatic, I hope this undertaking conjures up you to construct your individual recommender system!
Let me know your ideas and strategies within the feedback beneath.
