56 ROUND TABLE AI
In this month’s round table our experts discuss the opportunities for artificial intelligence and machine learning in the media industry
veryone is talking about AI, but not many people know what it really is – or is capable of. We got together some of our FEED Round Table experts to get to the bottom of the tech everyone is talking about. FEED: What are some of the most interesting media industry applications for artificial intelligence and machine learning right now? BHAVESH VAGHELA, PAYWIZARD: For us, the most interesting AI applications are focused around employing advanced predictive models based on subscriber data, which can provide pay-TV operators with a deeper understanding of their subscribers than ever before. For operators trying to provide the best experience for their subscribers on this journey, embracing AI can only improve that quality of engagement. Deeper AI capabilities are increasingly supporting operators to understand customer behaviours through modelling various data sources and historical activity, incorporating viewership, payment, subscription or even pay-per- view consumption. This in turn gives an operator the ability to understand their customer mix based on these key decision moments, and suggest the right action at the right time to reduce churn and improve customer experience. STUART COLEMAN, SOUTHAMPTON DATA SCIENCE ACADEMY: There is a huge amount of hype around AI and machine learning and like all hype curves it can be a minefield to wade through the real-world use case and those applications
which will make a difference to consumers and businesses. I may be oversimplifying, but the opportunity to apply AI and ML related tools and techniques fits neatly into three buckets. First is personalisation, which makes you or our clients think and feel like we have really taken time to prepare something unique and special. Imagine chatbot services online helping you engage with media services - content recommended just for you based on what you watch or engage with. Then there’s speed, with machines automating stuff humans typically do and/ or doing stuff we can do with less error. Think of curating and then text mining huge numbers of reviews from multiple channels and platforms. Or turning waves of unstructured data into structured information to better inform your marketing teams about your customers and services Finally there’s analytics. Finding new insights, trends and patterns. Marketing teams have always done this but the techniques, approaches and models which can be built and trained now go way beyond traditional customer-based segmentation. ANDREAS JACOBI, MAKE.TV: A I is helping to scale the huge task of acquiring, moderating, processing and producing massive amounts of video. For example, it will have a huge impact on how events are covered live because it will help identify the best content from a huge array of real-time sources much more quickly than human operators ever could without help. An interesting impact of this is that it will help broadcasters compete with social
I THINK AT THE MOMENT THERE’S A RUSH TO CALL LOTS OF THINGS AI media for audiences by making sure they can access the best content as it happens. Another application is in creating hyper-personalised viewing experiences. AI will make it much more likely that broadcasters will create personalised, generated live and on-demand content for different viewers. Again, this will help drive deeper audience engagement. AI will also help broadcasters monetise their libraries and save money reshooting. This is because AI will dive into libraries to highlight the most relevant content based on themes, stories, geography and people more quickly than traditional searches. dynamic channels comprised of professionally-produced and user-
feedzine feed.zine feedmagazine.tv
Powered by FlippingBook