Dynamic Invitational Content Selection via User Attention Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing media streaming services face suboptimal user engagement with invitational content due to arbitrary decisions on audio or video advertisement types, leading to decreased user experience and unsatisfied advertisers, as current methods do not effectively predict user attention levels to select the most appropriate advertisement type.
Innovation Solution
The system predicts user attention levels using client device data, such as screen lock state, device type, and ambient light, to select between audio and video invitational content items, prioritizing the type that aligns with the user's engagement level, and overrides based on sponsor or campaign preferences and environmental factors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If arbitrary decision mechanism is used to select advertisement type, then implementation is simple, but user engagement and advertising effectiveness deteriorate
Solution Approach 1:
The system changes the parameter of advertisement type selection from arbitrary to data-driven by analyzing multiple client device data parameters (screen lock state, device type, ambient light, device orientation, media station type) to predict user attention level and dynamically select between audio and video invitational content items, thereby improving user engagement while maintaining operational simplicity through automated prediction rules
2Productivity
If video advertisement is presented when user is not visually attentive, then advertising revenue increases, but user experience deteriorates
Solution Approach 1:
The system implements feedback by continuously monitoring client device data (screen lock state, device type, ambient light, device orientation, media station type) to predict user attention level, using this prediction feedback to dynamically select the most appropriate advertisement type (audio or video) that matches user engagement state, thereby maximizing advertising effectiveness while maintaining positive user experience
Solution Approach 2:
The system changes the advertisement presentation parameter from static to dynamic by adjusting the type of invitational content (audio vs. video) based on predicted user attention level derived from multiple device parameters, ensuring that advertisements are presented in a format that aligns with user engagement state
3Object-affected harmful factors
If audio advertisement is presented when user is visually attentive, then user experience is maintained, but advertising effectiveness is reduced
Solution Approach 1:
The system uses feedback from real-time analysis of client device data (screen lock state, device type, ambient light, device orientation, media station type) to predict user attention level, and adjusts advertisement type accordingly - presenting video advertisements when visual attention is detected and audio advertisements when visual attention is not present, thereby optimizing advertising effectiveness while maintaining user experience
Solution Approach 2:
The system makes the advertisement presentation dynamic by continuously adapting the invitational content type (audio or video) based on changing user engagement states predicted from device parameters, rather than using a fixed approach, ensuring optimal advertising effectiveness matched to real-time user behavior
Data Source
AI summary
A media channel can include a mix of media items and invitational content items. At some point during the playback of the media channel an invitational content item can be presented. In some cases, the invitational content items eligible for presentation can be of differing types, such as video and audio. In can be advantageous to restrict presentation of video invitational content items to times when a user is likely to view the screen of the client device during playback of the invitational content item. To accomplish this one or more heuristics or rules can be applied to client device data to predict a user attention level. The user attention level can then be correlated to an invitational content item type, which can then be used to select an invitational content item for playback.