Document Types
Paper Presentation
School Code
N/A
School Name
De La Salle University Integrated School
Abstract/Executive Summary
“Over the Top” platforms, or OTT platforms, are where movies and TV shows can be watched. The main focus of the research is the recommendation system of OTT platforms, studying its mechanisms. The researchers also aim to identify relevant features that are most useful to a recommendation system. The researchers conducted data preprocessing such as the one hot encoding method. Cosine similarity was employed as the foundational algorithm for the recommendation system. Upon generating several recommendations using different sets of features, the most relevant ones were determined through a survey. By utilizing the cosine similarity algorithm, the research aims to improve the OTT platform recommendation system. This study also seeks to gather data sets using standard pre-processing methods and identify the features that will result in the best recommendations when using the cosine similarity algorithm. The researchers compared different data sets and similarity scores based on various features. The researchers found that the data set with all gathered features had the highest level of similarity and is likely to be used in the recommendation system.
Keywords
OTT platforms; recommendation systems; data preprocessing; cosine similarity; feature selection
Research Theme (for Paper Presentation and Poster Presentation submissions only)
Computer and Software Technology, and Robotics (CSR)
Initial Consent for Publication
yes
Start Date
29-6-2023 10:30 AM
End Date
29-6-2023 12:00 PM
An Exploratory Study of OTT Platform Movie Recommendation using Cosine Similarity
“Over the Top” platforms, or OTT platforms, are where movies and TV shows can be watched. The main focus of the research is the recommendation system of OTT platforms, studying its mechanisms. The researchers also aim to identify relevant features that are most useful to a recommendation system. The researchers conducted data preprocessing such as the one hot encoding method. Cosine similarity was employed as the foundational algorithm for the recommendation system. Upon generating several recommendations using different sets of features, the most relevant ones were determined through a survey. By utilizing the cosine similarity algorithm, the research aims to improve the OTT platform recommendation system. This study also seeks to gather data sets using standard pre-processing methods and identify the features that will result in the best recommendations when using the cosine similarity algorithm. The researchers compared different data sets and similarity scores based on various features. The researchers found that the data set with all gathered features had the highest level of similarity and is likely to be used in the recommendation system.
