James' Portfolio

For the last week we went over the code again to make sure there are no errors in any way shape or form. Then we finished up by making the slides for the presentation. This was by far one of the most difficult projects that I have made in my days of Binus, but overall it was do able and pretty proud of the product that came out of it. It was a great learning experience in machine learning. The last thing we did was upload the code to GitHub and made the demonstration video that was required for us to make.

This week we are going to touch up on the code that was previously made the week before. In order to make it more readable and easier to understand we changed things around so that we can explain it better to the class. The code that we made was only able to filter based on movie titles, but we felt that was not enough so we decided to add some other features to it. The thing that we added was that you could base your filtering on genres and directors to give the program more variation rather than just filtering by titles. After completing this step all that is left to do is do prepare the demonstration video, the presentation slides, and uploading everything to GitHub.

We have finally found a complete data sets that consist of many movies and its appropriate datas that are needed for the project. We started this week by watching over the video again that will help us to bring this project to life, followed by brushing up on some python skills. Then we finally started to code using Colab by Google so that all of the libraries that are used are already there. No need for further installations which is really nice. We found a great source that will help us for this project and later will be cited in our presentation. We got the content based filtering to function with minimal functionalities, but in the following week we will continue to improve and if possible add our own features in order to make it more interesting.

This week our group met up to do some research in finding a well documented movie recommendation program to see what types of filtering or algorithms that are used. It was found that the most commonly used filtering system for these type of programs are popularity based filtering, content based filtering, and collaborative filtering. Which is great for us because content based filtering is what we planned on doing so we are finding some sources in order to implement that type of filtering into our project. Watching some YouTube videos as well in order to understand more on this type of filtering and Also finding a large data set with lists of movies and its appropriate data.

After the mid exam our group conducted a research between types of filtering in order to make our movie recommendation program. We as a group, decided to do a content based filtering program. In this type of filtering the user will be asked to input a movie title, genre, or director. Then it should list out movies based on the user’s input. But the goal for this week is to find out more about this type of filtering and find a way to implement it into our project and make it come to life.