Expert Identification Indexing in Enterprise Social Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems for identifying experts on arbitrary topics in enterprise social networks face challenges, such as low profile updates and the need for accessing private information, leading to delayed and unreliable results.
Innovation Solution
A method that indexes publicly available content within an enterprise social network, using activity signals like likes, views, and bookmarks to rank search results, providing immediate and secure identification of experts without accessing private data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing systems access private information to identify experts, then identification accuracy may improve, but security risks and system complexity increase
Solution Approach 1:
The patent extracts only the necessary public information (profile data, activity signals, content contributions) from the social network, excluding private information. This selective extraction maintains identification accuracy while eliminating security risks associated with accessing private data.
Solution Approach 2:
The system introduces an intermediary indexing layer that processes and aggregates public activity signals without directly accessing private user data. This intermediary structure enables accurate expert identification while maintaining a security boundary that prevents harmful access to private information.
2Reliability
If existing systems wait for profile updates to identify experts, then data reliability improves, but response time deteriorates
Solution Approach 1:
The system performs preliminary indexing of public profile data and activity signals in advance, creating a ready-to-query knowledge base. This preliminary action ensures data reliability through proper indexing while enabling immediate expert identification without waiting for profile updates, thus reducing response time.
Solution Approach 2:
The indexing system continuously monitors and updates public activity signals (likes, views, bookmarks, content contributions) in real-time, maintaining reliable data without requiring users to manually update profiles. This continuous automatic updating eliminates response delays while ensuring data reliability.
3Adaptability or versatility
If the system indexes all content to enable arbitrary topic searches, then search completeness improves, but system complexity and resource consumption increase
Solution Approach 1:
The patent segments the indexing process by creating separate indexes for different types of content and activity signals (profiles, posts, likes, views, bookmarks). This segmentation enables complete arbitrary topic searches across all content types while reducing overall system complexity by organizing data into manageable, specialized index structures.
4Measurement precision
If the system uses multiple activity signals to rank results, then identification precision improves, but computational requirements increase
Solution Approach 1:
The system changes parameters by assigning different weights to various activity signals (likes, views, bookmarks, content contributions) based on their relevance to expert identification. This weighted approach improves identification precision by emphasizing more meaningful signals while reducing computational energy through selective processing rather than treating all signals equally.
Data Source
AI summary
A system and method for supporting identifying experts on arbitrary topics in an enterprise social network. An exemplary method can receive, at a node of a social network application, a plurality of content. The method can store the plurality of content in a database associated with the social network application. The method can index the plurality of content, resulting in an index of content, wherein the index of content includes a plurality of activity signals and a plurality of security values. The method can receive a request for a search for at least one expert on an arbitrary topic. The method can search the index of content for the at least one expert on the arbitrary topic, resulting in a plurality of search results. The method can next assign a weight for each of the plurality of search results based at least on the plurality of activity signals.