https://view-wise.streamlit.app/
ViewWise is a recommendation system project that suggests TV shows based on cosine similarity between their metadata. By analyzing aggregated textual data of TV shows, the system provides users with personalized recommendations from a curated list of popular shows.
ViewWise aims to enhance recommendation accuracy by leveraging aggregated text descriptions and ratings for TV shows. The system matches users with shows similar to their selection, presenting alternatives that align with their viewing preferences. This enables users to discover content beyond their usual picks while maintaining high relevance to their interests.
In the competitive streaming industry, personalized recommendations play a pivotal role in retaining user engagement. ViewWise provides a deeper analysis of TV shows by factoring in not only genre or cast but also audience reviews, ratings, and plot similarities. It goes beyond traditional methods by offering:
-
More relevant show recommendations based on detailed metadata.
-
A richer discovery experience by connecting users to hidden gems that fit their taste.
-
Enhanced user retention through precise personalization based on content similarity.
-
Building a recommendation system experimenting with both Count and TFIDF Vectorizer.
-
Handling processing of text data, such as aggregating text, handling special characters and stopwords.
-
Web scraping using Selenium and Beautiful Soup, managing dynamic content, and interacting with HTML elements effectively.
- Last Updated: 5/9/2024 (Updates Yearly)
- (Include TV Shows IMDB Rating > 7.5, No. Ratings > 50,000)
- By: Hong Kai