Guided Query Graph Navigation for Context-Specific Search
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems fail to provide sufficient, context-specific information for professionals like architects or developers, as search engines yield irrelevant results, while topic-specific websites require manual search across multiple sources, and machine learning lacks relevant data for specialized domains.
Innovation Solution
A computer-implemented guidance system using a graph database and context memory to guide users through relevant information by selecting guidance root nodes, grouping options, and queuing context-relevant suggestions, without relying on machine learning.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If search engines are used to find information, then a wide range of information is accessible, but the results are generic and irrelevant to specific professional contexts
Solution Approach 1:
The system segments information retrieval into specialized domains by creating domain-specific knowledge graphs that divide the vast information space into manageable, context-relevant segments. This allows the system to retrieve precisely targeted information for specific professional contexts rather than returning generic search results across all domains.
Solution Approach 2:
The patent introduces an intermediary layer between the user and information sources through a chatbot interface that maintains conversation context and mediates information retrieval. This intermediary processes user queries through multiple turns, maintaining contextual understanding and translating vague requests into precise information needs, thereby bridging the gap between generic search capabilities and specific information requirements.
2Loss of information
If topic-specific websites are used to find specialized information, then relevant domain information is available, but manual searching across multiple sources is required
Solution Approach 1:
The system merges multiple domain-specific information sources into a unified knowledge graph that integrates data from various specialized websites and databases. This consolidation allows the system to retrieve comprehensive domain-specific information from multiple sources through a single query, eliminating the need for manual searching across multiple separate websites.
Solution Approach 2:
The system performs preliminary action by pre-processing and structuring domain-specific information into organized knowledge graphs before queries are submitted. Information from multiple sources is预先 organized, tagged, and interconnected in advance, enabling rapid retrieval without requiring manual searching at query time. The knowledge graphs are built and maintained proactively rather than reactively.
3Extent of automation
If machine learning recommendation systems are used, then automated suggestions are provided, but insufficient domain-specific data is available for training
Solution Approach 1:
The patent replaces the statistical machine learning approach with a rule-based expert system that uses manually crafted knowledge graphs and domain expertise. Instead of relying on large volumes of training data to learn patterns, the system uses explicitly defined rules, ontologies, and structured knowledge representations that capture domain-specific relationships, enabling automated recommendations with limited data.
Solution Approach 2:
The system changes the fundamental parameters of the recommendation approach by shifting from data-driven statistical methods to knowledge-driven rule-based methods. This parameter change involves transitioning from learning patterns through extensive training to applying pre-established domain knowledge and rules, thereby enabling effective automation even with limited quantitative data available.
4Adaptability or versatility
If catalogues with filtering abilities are used, then multiple options can be explored, but too many options are presented at once overwhelming the user
Solution Approach 1:
The system applies dynamics by making the information presentation adaptive and conversational rather than static and comprehensive. The chatbot dynamically adjusts the number and type of options presented based on the conversation context, user responses, and accumulated knowledge. Information is revealed incrementally through multiple turns rather than all at once, maintaining simplicity while enabling thorough exploration.
Solution Approach 2:
The system uses periodic action by presenting information in structured phases or turns rather than continuously. The chatbot engages users in a sequence of question-answer cycles, where each turn builds on previous responses and gradually narrows down options. This periodic interaction pattern allows comprehensive option exploration while maintaining user interface simplicity through paced, manageable information delivery.
Data Source
AI summary
Computer system (GUID) and computer-implemented method for a guidance session comprising measures like receive a user's request (req), select a guidance root node (r1, r2, rn) based on the user's request (req) by accessing a graph data base, wherein the graph data base provides guidance options (o1, . . . on) with links (l), group (grp) in an iterative manner guidance options (o) by type (ot) and put them into a queue (qu), dequeue (dequ) each option group (og2/og3) and check each option's relevance (rel) in the context, provide (show) the option (o) to the user (u) and provide a first guiding question, store user specific interactions history in the context memory and/or perform option-specific actions.