A demo movie recommendation system
\(\quad\) \(\quad\)
The prosperity of Internet industry enabled massive knowledge and entertainments to be accessible at one’s fingertips, making people no longer in thirst for feeds. However, massive information also trapped people into another dilemma: we got numerous choices for our limited attentions. Therefore, many Internet companies, such as Youtube, Tiktok and Bilibili, has equipped their website with personalized recommendation service to attract users’ attentions as much as possible. Curious about how the recommendation works, we learned and implemented several recommendation algorithms, through applying some of which we built our own movie recommendation system upon movielens dataset.
\(\quad\) \(\quad\)
This website introduced algorithms and codes we used to built a demo movie recommendation upon the movielens dataset. The dataset described rating measured in the 5-star format and free-text tagging activity from the a real film websites. When exploring the website, you will find the Exploratory Analyses, Differentiable Movies, Recommendation Algorithms and a Shinyapp as our recommendation system.
\(\quad\)
In the Exploratory Analyses tab, you will find:
Brief overview of the dataset
In-depth analyses of the dataset, through statistical tests and visualization
\(\quad\)
In the Differentiable Movies tab, you will find:
Answer to the question: what movies should be provided to new users for rating to start recommendation?
Core ideas of the adaptive bootstrapping algorithm
Implementation of this algorithm in R
\(\quad\)
In the Recommendation Algorithm tab, you will find the concepts, implementation and test results for three recommendation algorithms:
\(\quad\)
The shinyapp tab will lead you to the demo recommendation system we built.
You can also see our proposal here.
\(\quad\) \(\quad\)