Social Media Post Quality Estimation Using Member Heterogeneity

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing social media post presentation methods are overwhelmed by the sheer volume of content, and content-based quality estimation fails to consider contextual relevance, leading to inappropriate post display in certain groups.

Innovation Solution

A content-agnostic quality estimation approach using a 'wisdom of the crowd' method, where the diversity of member skills inversely correlates with post specificity, computes a heterogeneity score based on member attributes like skills and work titles to determine post quality and relevance, aiding in contextual topic modeling.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If content-based quality estimation is used to filter social media posts, then post quality is improved, but contextual relevance is lost

Engineering Contradiction:

Improvepost qualityVSAvoidcontextual relevance

Core Design Contradiction:

Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The patent transitions from evaluating posts in a single dimension (content quality) to multiple dimensions by incorporating member attribute heterogeneity. This adds a contextual dimension to quality estimation, allowing the system to assess both content quality and contextual appropriateness simultaneously, thereby resolving the contradiction between content-based quality and contextual relevance

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

Solution Approach 2:

The patent changes the evaluation parameters from solely content-based metrics to include member attribute-based metrics (heterogeneity scores). By introducing new parameters that capture the diversity of member skills, education, and demographics, the system can evaluate posts differently based on group context, thus maintaining quality assessment while adapting to contextual relevance

Inventive Principle:

Principle #35Parameter changes

2Quantity of substance

If all social media posts from linked members are displayed in a feed, then content completeness is improved, but user experience deteriorates due to overwhelming volume

Engineering Contradiction:

Improvecontent completenessVSAvoiduser experience

Core Design Contradiction:

Quantity of substanceVSEase of operation

Solution Approach 1:

The patent applies local quality by differentiating post evaluation based on the specific group context. Instead of applying a uniform quality threshold across all posts, the system adjusts quality standards according to the heterogeneity characteristics of each group, allowing relevant posts to be displayed while filtering out inappropriate ones, thus maintaining completeness without overwhelming users

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The system uses heterogeneity scores as feedback mechanisms to dynamically adjust post display decisions. By continuously evaluating the match between post content and group member attribute profiles, the system provides feedback-driven filtering that maintains content completeness while enhancing user experience through relevant post selection

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS11288334B2Content agnostic quality estimation and topic modeling

Publication Date: 2022.03.29 MICROSOFT TECHNOLOGY LICENSING LLC

AI summary

In an example embodiment, a content agnostic approach to quality of assessment of social media posts or other items or content in an online network, such as a social networking service, is utilized. Specifically, information about members contained in member profiles may be used to derive meaningful insights about posts they interact with collectively.