Entity Graph Grouping for Intuitive Search Presentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current search systems do not effectively provide intuitive and compact information about groups of related entities to users, failing to rank and present these relationships in a user-friendly manner.
Innovation Solution
A system that identifies related entities through an entity graph by determining paths with the same relationship type, ranks them based on popularity and relationship strength, and generates presentations that include text descriptions of these relationships, allowing for more comprehensive and accessible information about entities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If the system provides comprehensive information about all related entities, then the information completeness is improved, but the presentation becomes complex and difficult to navigate
Solution Approach 1:
The system segments related entities into distinct groups based on relationship types (e.g., family members, colleagues, works). Each group is presented separately with a clear heading, allowing users to comprehend comprehensive information without being overwhelmed by a single undifferentiated list. This segmentation resolves the contradiction by organizing complete information into manageable, navigable sections.
Solution Approach 2:
The system adds a dimensional organization layer by grouping entities along the relationship type dimension. Instead of presenting all related entities in a single flat list, entities are arranged in multiple dimensions: first by relationship category, then by individual entity within each category. This dimensional approach maintains information completeness while reducing presentation complexity through structured organization.
2Ease of operation
If the system ranks and presents only the most popular relationships, then the ease of operation is improved, but the information completeness deteriorates
Solution Approach 1:
The system implements dynamic presentation where the most popular or relevant relationship groups are displayed prominently first, but additional groups remain accessible through expansion options or pagination. This dynamic approach allows the interface to adapt to user needs: providing immediate ease of operation by highlighting top relationships while preserving information completeness by making other relationships accessible upon user request.
3Loss of information
If the system provides detailed information about each related entity, then the information completeness is improved, but the screen space requirement increases
Solution Approach 1:
The system uses a nested presentation structure where relationship group headings contain or are associated with their member entities. Each group serves as a container that organizes multiple entities, allowing hierarchical display where summary information appears at the group level and detailed entity information is nested within. This nesting enables complete information presentation while efficiently utilizing screen space through hierarchical organization.
4Ease of operation
If the system groups entities by relationship type, then the intuitiveness is improved, but the system complexity increases
Solution Approach 1:
The system automatically determines relationship types and groups entities without requiring manual configuration. The entity graph structure and relationship data enable the system to self-organize entities into intuitive groups based on their inherent relationships. This self-service approach improves intuitiveness for users while avoiding the complexity of manual group configuration, as the system autonomously performs the grouping based on pre-existing relationship data.
Data Source
AI summary
Methods, systems, and apparatus, including computer programs encoded on computer storage media, for grouping related entities. One of the methods includes receiving a search request that includes query text that identifies a first entity. A first entity node in an entity graph that represents the first entity is identified. A highest ranked first group of related entity nodes from a ranking of a plurality of groups of related entity nodes is identified, wherein each related entity node in the first group is connected to the first entity node by a respective path from the related entity node through one or more links in an entity graph. A presentation of information associated with one or more entity nodes in the first group of related entity nodes is provided.