Our New Data Science Competition To Predict Video Views

By building an algorithm to predict how many views each video on their platform is likely to receive, a popular video hosting website will be able to optimize ads, video recommendations, and advice for creators.

We launched our first data science competition of 2021 last week, called the Video Popularity Prediction Challenge. Running until March 31, this challenge asks participants to predict how many views videos are likely to receive based on a unique dataset provided by Japanese startup DataGateway.

This dataset includes data from 3,000 videos spanning a nine-year period from 2011 to 2020. Here is a breakdown of the columns you’ll find in the dataset on the competition website:

Other datasets are also available on our website to help the data scientists who join to more accurately predict these view numbers, including:

• Thumbnail image data
• Vectorized title data
• Vectorized description text data

For more information and to download these datasets to get started, sign up on the competition’s landing page here.

Note: This competition is now closed. You can find a list of bitgrit’s other AI competitions on our website here.