Social Graph Interest Profiling Across Multiple Social Accounts
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional approaches face challenges in identifying the same person across multiple social media accounts due to the use of different names or identifiers and determining a person's interests in a concise manner for decision-making processes such as hiring or group membership.
Innovation Solution
A method involving network mapping, image processing, and machine learning techniques to aggregate social media interactions, identify clusters of interests, and determine a person's risks based on their social network activities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional approaches are used to identify persons across social media accounts, then the process is simple, but the accuracy is poor due to different names or identifiers
Solution Approach 1:
The patent introduces a social graph as an intermediary structure to connect different social media accounts. By mapping accounts to a unified social graph with unique identifiers, the system can accurately identify the same person across different platforms without requiring direct matching of names or identifiers, thus improving identification accuracy while managing complexity through structured representation.
Solution Approach 2:
The patent transitions from a flat, direct matching approach to a multi-dimensional social graph structure. By adding the dimension of social relationships and network context, the system can infer person identity through their connections rather than direct identifier matching, enabling accurate identification even when names or handles differ across platforms.
2Loss of information
If detailed analysis of social media interactions is performed, then the completeness of interest identification is improved, but the processing time and resources increase
Solution Approach 1:
The patent extracts and aggregates interest information from diverse social media interactions into a unified, normalized representation. By extracting only the relevant interest signals and consolidating them across platforms, the system maintains complete interest identification while reducing processing time through selective extraction rather than exhaustive analysis of all interactions.
Solution Approach 2:
The patent merges interest data from multiple social media sources into a single consolidated profile. By combining and aggregating interest information across different platforms and interaction types, the system achieves comprehensive interest identification without requiring separate processing of each source, thereby reducing overall processing time through consolidation.
3Measurement precision
If image processing is performed for all social media content, then the accuracy of interest determination is improved, but the computational resources required increase significantly
Solution Approach 1:
The patent applies image processing selectively rather than universally. By determining whether image analysis is needed based on the specific content type and context, the system maintains high accuracy for cases where images are informative while avoiding unnecessary computational resources for text-based or already-processed content, thus optimizing the balance between accuracy and resource consumption.
Data Source
AI summary
Systems, devices, and methods for identifying a person's interests based at least in part on their interactions with social networks, where “interests” may refer to a person's hobbies, opinions, behaviors, goals, memberships, or activities, as non-limiting examples. The person's interactions with social networks may include postings of comments, articles, or images, indicating a like or dislike for a posting of another person, “following” an account, performing a search, or the type of networks or social media they engage with (which may be associated with a social or political goal, as a non-limiting example).