Comment Ranking via Interaction-to-Impression Ratio

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Social networking systems face challenges in optimally ranking comments to posted content due to the 'early engagement effect,' where earlier comments receive artificially inflated rankings, while more relevant subsequent comments are pushed down, despite higher interaction-to-impression ratios indicating greater user interest.

Innovation Solution

Implementing a comment ranking system that utilizes an interaction-to-impression ratio, combined with relevance determination and user analysis, to rank comments based on their relevance to the original content, user interactions, and user history, thereby discounting early interactions and prioritizing more relevant comments.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If comments are ranked based on total number of interactions, then comments with more interactions receive higher rankings, but this creates artificial inflation for early comments and pushes down more relevant subsequent comments

Engineering Contradiction:
Improvecomment ranking accuracyVSAvoiduser interest signal
Core Design Contradiction:
ReliabilityVSLoss of information

Solution Approach 1:

The patent changes the ranking parameter from absolute interaction count to interaction-to-impression ratio. This parameter transformation normalizes the interaction data by dividing by the number of impressions, thereby removing the artificial inflation effect and allowing subsequent comments with higher engagement rates to be ranked appropriately.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent introduces a new dimension to the ranking system by incorporating impression data as a divisor. This transforms the single-dimensional interaction count metric into a two-dimensional interaction-to-impression ratio metric, enabling the system to distinguish between comments with high absolute interaction counts and comments with high engagement rates.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Productivity

If earlier comments are ranked higher due to early engagement, then user feedback is captured quickly, but relevant later comments are suppressed despite showing higher interaction-to-impression ratios

Engineering Contradiction:
Improvecomment engagement speedVSAvoidcomment relevance ranking
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

By changing the ranking parameter to interaction-to-impression ratio, the system maintains the ability to quickly capture user feedback while simultaneously enabling relevant later comments to surface. The ratio metric allows the system to recognize high engagement rates regardless of comment timing.

Inventive Principle:
Principle #35Parameter changes

3Quantity of substance

If interaction rating is used for ranking, then user interactions are rewarded, but the rating must be discounted based on interaction-to-impression ratio to prevent early engagement inflation

Engineering Contradiction:
Improveinteraction volumeVSAvoidranking accuracy
Core Design Contradiction:
Quantity of substanceVSReliability

Solution Approach 1:

The patent applies parameter transformation by converting the interaction rating into an interaction-to-impression ratio. This mathematical transformation (dividing interaction rating by impression count) automatically discounts early engagement inflation while preserving the signal of genuine user interest, enabling accurate ranking without manual adjustment.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentUS10419383B2Systems and methods for ranking comments based on interaction-to-impression ratio
Publication Date: 2019.09.17 META PLATFORMS INC
  • US10419383B2 patent drawing
  • US10419383B2 patent drawing
  • US10419383B2 patent drawing

AI summary

Systems, methods, and non-transitory computer-readable media can receive a comment to a posted content item. One or more user interactions with the comment are received. An interaction-to-impression ratio for the comment is determined. The comment is ranked among a plurality of comments based on the interaction-to-impression ratio.