Whether it’s Youtube, Reels, or TikTok, most video sharing applications have a tab that lists their trending videos. How do you find the most “happening” and “engaging” one in a video sharing platform? Five hundred hours of video are uploaded to YouTube every minute, and people watch over 1 billion hours of YouTube videos a day, more than Netflix and Facebook combined. With hundreds and thousands of new videos getting uploaded in such an application, you need to have a seamless algorithm to find the most popular ones.
App development companies and dedicated teams can help you in narrowing down the right algorithm. Here at Perfomatix, being one of the top app development companies – we keep innovating with our technical team. Our frontend software engineer – Hanz Haneefa found a simple yet effective solution to find trending videos in a video sharing application.
Factors to consider to find the trending videos
When you consider a video sharing platform, for finding out the trending videos – one need to take into account various features like the number of views, the upload time, number of likes and or dislikes, comments, location of video uploaded, video category, and overall engagement – shares/social media shares/ hashtags, etc. These numbers will guide you towards a general understanding of what’s the most popular video.
Still, you wouldn’t get a clear idea to unify these numbers and find a standard “trending value”. Most importantly, the algorithm to find the trending video should consider the upload time because a 500-day old video garnering many views and comments shouldn’t be trending over a week old video has a lesser number of views. Keeping these factors in mind, our developers decided to create a simple algorithm that will efficiently calculate the “trending factor” of uploaded videos.
Trending value calculator algorithm
For instance, in a new algorithm developed by our team – we considered the average number of views in a particular time interval, and we recorded this data continuously. Some of the other factors like the number of votes (likes and dislikes) & comments were normalized to find an average value.
This normalized value for views, votes, and comments was assigned a weighted average, resulting in the trending value. According to various other factors depending on the application – like location, category, and more this data can be further customized to find trending videos in various specifications. At the core, our simple algorithm developed by Software Engineer Hanz Haneefa helped simplify the task of finding out the trending videos.
Check out the algorithm in Github here.
Need help? Perfomatix | App Development Company
We are Perfomatix, one of the top mobile app development companies. We provide mobile app development services in building highly scalable android and iOS apps, and APIs. We also have strong expertise in IoT apps, Virtual Reality apps, and Augmented Reality apps. Get in touch with us to find out how we can help in shaping your disruptive idea into a prototype, MVP, and finally into a killer product. Visit our success stories section to find out more about some of the startups which made it big with us.
