Cross-Team Post Ranking Using Communication Graph Latency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current recommendation systems fail to optimize cross-team information flow and consider temporal aspects of communication, leading to obsolete recommendations that do not enhance communication across teams and reduce productivity.
Innovation Solution
A recommendation system that utilizes an algorithm measuring information latency and total information metrics to optimize cross-team communication by recommending posts that consider the temporal flow of information and relevance, using a communication knowledge network graph to rank posts based on their impact on information flow.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If traditional recommendation systems recommend posts based on user-by-user relevance, then individual user relevance is improved, but cross-team information flow is not promoted
Solution Approach 1:
The patent transitions from a single-dimension user-by-user relevance approach to a multi-dimensional approach that incorporates team-level information flow metrics. By adding the dimension of team-to-team information propagation to the recommendation criteria, the system simultaneously optimizes both individual user relevance and organizational cross-team information flow.
Solution Approach 2:
The recommendation system is enhanced to serve multiple functions: it not only recommends posts relevant to individual users but also promotes cross-team information flow and reduces information latency across the organization. This multi-functional approach allows a single recommendation system to address both user-level and organization-level objectives.
2Device complexity
If recommendation systems do not consider temporal aspects of communication, then system simplicity is maintained, but information latency increases and recommendations become obsolete
Solution Approach 1:
The system pre-calculates and stores team communication graphs and information latency metrics before generating recommendations. By preparing this temporal context information in advance, the system can quickly assess the timeliness and potential impact of recommended posts without adding significant complexity to the real-time recommendation process.
Solution Approach 2:
The patent replaces simple relevance-based filtering with a more sophisticated temporal-aware recommendation mechanism that uses communication graphs and latency metrics. This substitution introduces temporal dimensionality to the recommendation system, enabling it to account for information freshness and propagation dynamics while maintaining computational efficiency through graph-based representations.
3Ease of operation
If recommendation systems filter out content to manage information overload, then user burden is reduced, but information dissemination efficiency decreases
Solution Approach 1:
The system uses feedback from the communication knowledge graph about how information flows between teams to guide recommendations. By analyzing actual communication patterns and information propagation paths, the system can prioritize posts that are most likely to improve cross-team information flow, thereby reducing user burden while enhancing dissemination efficiency through data-driven prioritization.
Solution Approach 2:
The recommendation system changes the parameters used for filtering and prioritizing content from simple user-level relevance scores to include team-level information flow metrics and latency measurements. This parameter transformation enables the system to identify and promote posts that will most effectively improve cross-team information flow while still managing user information overload through selective prioritization.
Data Source
AI summary
A system and method and for optimizing cross-team information flow in a communication network includes receiving, from a communication application, via a network, a plurality of candidate post items for display to a first user of an organization, each candidate post item being a post item published by another user of the organization and being a post that is accessible to the first user. A communication knowledge network graph which represents communication events that have occurred between users of the organization is then generated where each communication event is represented by a first node that represents a sender, a second node that represents a receiver and an edge that represents the communication event from the sender to the receiver. For each one of the candidate post items, a value of a total information metric for the communication knowledge network is estimated if one of the plurality of candidate posts is viewed by the first user and the plurality of candidate post items are ranked based on the estimated total information metric before transmitting recommendation data to the communication application for recommending the plurality of candidate posts to the user based on the ranking.