News Stream Item Prioritization via Normalized Dwell Times
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing news personalization systems face challenges in effectively prioritizing items of different media types, such as news articles, videos, and slideshows, due to varying user engagement metrics and lack of contextual information, particularly for images, which complicates the creation of a customized news stream that maintains user interest.
Innovation Solution
The method involves calculating dwell times for each media type, normalizing these metrics to account for differences in media characteristics, and using z-scores to prioritize items in a unified news stream, ensuring that items from diverse categories are ranked based on user engagement signals, thereby creating a personalized and engaging news feed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional news personalization systems prioritize items based on simple click metrics, then the system is easy to implement, but it cannot effectively compare and rank items of different media types due to varying user engagement patterns
Solution Approach 1:
The patent transforms the engagement measurement parameter by introducing normalized dwell time metrics that account for media-type-specific viewing patterns. Instead of using raw click data, the system calculates average dwell times for each media type and normalizes individual item dwell times against these averages, enabling accurate comparison across different media formats while maintaining a unified ranking framework
Solution Approach 2:
The patent introduces normalized dwell time as an intermediary metric that bridges the gap between different media types. This intermediary measurement standardizes user engagement data across articles, videos, and other media formats, allowing the system to rank diverse items fairly without requiring complex multi-criteria decision-making mechanisms
2Adaptability or versatility
If the news stream includes diverse media types to enhance user interest, then user engagement improves, but prioritizing items becomes more difficult due to lack of contextual information in certain media types
Solution Approach 1:
The patent shifts from content-based analysis to behavior-based measurement by using actual user dwell time data. Instead of attempting to analyze and extract meaning from diverse media types (which is particularly difficult for images and videos), the system measures how long users actually engage with each item, transforming the measurement parameter from content characteristics to user behavior metrics
Solution Approach 2:
The patent enables the system to self-adjust to different media types by automatically collecting and analyzing user interaction data. The system does not need to understand the content of each media type; instead, it lets user behavior data speak for itself, with dwell times naturally revealing engagement quality across all media formats without requiring manual content analysis
3Productivity
If the system uses raw dwell time metrics to prioritize items, then the prioritization process is simple, but it fails to account for differences in typical viewing durations across media types
Solution Approach 1:
The patent transforms the raw dwell time parameter into a normalized engagement metric. For each media type, the system calculates the average dwell time and then normalizes individual item dwell times against this average. This parameter transformation preserves the simplicity of using dwell time data while eliminating the bias introduced by inherent differences in viewing durations across media types
Solution Approach 2:
The patent creates equipotential conditions for comparing different media types by normalizing all dwell times to a common reference frame. By dividing each item's dwell time by its media type's average, the system equalizes the comparison baseline across articles, videos, and other formats, allowing fair ranking without requiring complex adjustment mechanisms
Data Source
AI summary
Methods, systems, and computer programs are presented for displaying a customized news stream. One method includes an operation for identifying dwell times spent by users while accessing a first plurality of items, each item belonging to one media type from a plurality of media types (e.g., news articles, videos, slide shows, etc.). In addition, the method includes operations for determining statistical parameters for each media type based on the identified dwell times, and for detecting a news corpus having a second plurality of items. A priority for each item in the news corpus is determined based on the media type of the item, the corresponding statistical parameters for the media type of the item, and the profile of a user. The news stream is sent to the user for presentation on a display, the news stream being sorted based on the priority of the items in the news corpus.