Human behaviour is very much driven by emotion. So it follows that emotions also influence our media consumption – such as in how we decide what we want to watch on video streaming platforms. There is currently a great deal of talk about sentiment, emotions, feelings and mood when it comes to analysing and curating streaming platform video content. Understanding the emotional effect of content supports improvements in both content recommendation and engagement as well as content development.
Viewers of streaming services often use their feelings to describe the content they find most relevant to their personal preferences and ask for recommendations in viewing groups for new things to watch, based on the emotional reaction caused by a series, movie, or other content they have just watched.
Viewers will also avoid some content because they will react to it in a negative way. If producers know how to analyse the content they publish more widely by looking at its emotional effects, they can offer more relevant recommendations and instead of spending time browsing a service, viewers can concentrate instead on watching the actual content.
The report [click here to access What are you in the mood for? ] provides an introduction to how metadata and Artificial Intelligence (AI) could be used to recommend content on a video streaming platform, based on the mood and feel it gives to the audience. It will also suggest ideas for the future in this field of emotional recommendations, using AI with humans in the loop. In addition to its opportunities and benefits, an emotion-based recommendation also raises considerations about ethics and risks. This article is based on expert interviews conducted in October 2022 and follow-up interviews in spring/summer 2023.
This report was written by Elina Yli-Ojanperä, a Publishing executive at the Finnish public media service company Yle. She specialises in working with digital content and platforms to provide an objective view from the customer’s perspective. Her expertise includes supporting digital content development, creating publishing plans, and understanding digital and streaming platform features and data analysis. Contact: firstname.lastname@example.org
This paper was produced as part of the Polis Newsroom Fellowship scheme supported by Yle. Polis is the international journalism think-tank at the London School of Economics and Political Science. This project was overseen by Polis director, Professor Charlie Beckett. The views in this paper reflect those of the author and are not necessarily those of Yle, Polis or LSE.
Click below to read the report PDF