Graph Inference Engine for Missing Knowledge Facts
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Relational models of knowledge, such as graph-based data stores, often lack essential facts and attributes, leading to incomplete search queries, as they may miss tens of millions of facts or contain incorrect relationships between nodes.
Innovation Solution
A computer system infers missing facts and attributes by analyzing related information in the graph, using a series of connections and edges to determine the absent information, and can search the internet for documents to verify and generate these missing facts, thereby enhancing query resolution.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If a graph-based data store is used to represent knowledge, then the system can store and organize a significant amount of facts, but the graph may be missing tens of millions of facts or have incorrect relationships between nodes
Solution Approach 1:
The patent introduces an inference engine as an intermediary component that mediates between the incomplete graph data and the user's query needs. This inference engine uses machine learning models to predict missing facts and relationships, effectively bridging the gap between the stored data and the required information without modifying the original graph structure.
Solution Approach 2:
The system performs preliminary inference actions by pre-computing and caching predicted missing facts and relationships before they are actually needed. The inference engine proactively identifies and fills in missing connections in the graph, so that when queries are executed, the data is already enhanced with inferred information, improving both completeness and query accuracy.
2Measurement precision
If the system infers missing facts by analyzing related information and searching the internet, then query resolution accuracy improves, but the processing time and system complexity increase
Solution Approach 1:
The system performs inference operations in advance and caches the results. When a query is received, the system first checks if the required information has already been inferred and stored. This preliminary action significantly reduces the processing time for subsequent queries, as the expensive inference operations are performed only once rather than repeatedly.
Solution Approach 2:
The inference engine applies different inference strategies and levels of detail to different parts of the graph based on local characteristics. For well-connected nodes with abundant information, less aggressive inference is applied, while for sparsely connected nodes, more intensive local inference is performed. This localized approach optimizes processing time by avoiding unnecessary inference operations throughout the entire graph.
3Adaptability or versatility
If multiple features and algorithms are used to infer missing facts, then the system's ability to resolve queries improves, but the device complexity increases
Solution Approach 1:
The inference system is segmented into multiple independent modules, each responsible for a specific type of inference task or a particular aspect of the data. This modular architecture allows the system to incorporate multiple features and algorithms without creating a monolithic complex system. Each module can be developed, tested, and maintained independently, and modules can be selectively activated based on query requirements.
Solution Approach 2:
The inference engine is designed as a universal platform that can handle multiple types of inference tasks through a common framework. The same core infrastructure supports various inference algorithms and features, allowing the system to adapt to different query types and data characteristics without requiring separate systems for each function. This multi-functionality reduces overall system complexity compared to having dedicated systems for each inference task.
Data Source
AI summary
In one example embodiment, a computer system includes at least one processor and a memory storing a data graph and instructions. The instructions, when executed by the at least one processor, cause the system to generate a template sentence based on a fact including a first node, a second node and a string, wherein the first node and the second node exist in the data graph and the string represents a fact that is absent from the data graph, search the internet for a document including the template sentence, and upon determining the internet includes the document with the template sentence, infer the fact by generating a series of connections between nodes and edges of the data graph that together with the first node and the second node are configured to represent the fact, the series of connections defining a path, in the data graph, from the first node to the second node.