Help with Building a Movie Recommendation Model Using KNIME

KNIME Analytics Platform
1

Hello everyone,

I’m relatively new to KNIME, but I’m excited to start building a simple model to predict whether a user will like a specific movie based on their past reviews. Initially, I wanted to create a model that could recommend the perfect movie for a group’s movie night, but I’ve realized that this might be too ambitious for my current time frame.

For this project, I plan to use the MovieLens dataset, which includes:

  • User data: Information such as user ID, age, gender, and occupation.
  • Movie data: Information about the movies, including movie ID, title, and genres (e.g., Action, Comedy, Drama).
  • Ratings data: Ratings provided by users, including user ID, movie ID, rating (1-5), and timestamp.

I would greatly appreciate any guidance or advice on how to approach this in KNIME!

Thanks in advance,
M

2

Hi @Menze and welcome to the forum.

We have a blog post available about this this exact use case and dataset: Movie Recommendations with Spark Collaborative Filtering | KNIME

The related workflow is available on the KNIME Community Hub here:

Admittedly it’s a few years old now, and there are multiple ways to approach the task, but maybe this helps give you an idea of how to get started.

1 Like
3

Hey,

Thanks for sharing your the post! I had seen it and it sounds really interesting, but it feels a bit advanced for where I’m currently at with KNIME. I’d like to start with something simpler to get the basics down. Could you help me figure out how to create a basic model as a starting point?

Thanks!

4

I think you could start with some of the examples you can find following this link:

02, 05 and potentially 09 should be relevant to what you want to achieve.

1 Like
5

Thank you for your help! I have been trying a bit but somehow I cannot seem to figure it out. I only get the model to predict whether a user is going to like a movie he has already rated. Obviously this is not what I want. I want the model to predict whether a user is going to like a movie he has not yet seen. Could you help me further? (please keep in mind I have absolutely no idea what I am doing)