Semantic Graph Question Generation for Data Exploration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Data analysts face challenges in determining the starting point for data analytics processes, as they spend significant time exploring data to discover meaningful relationships, and current solutions lack the ability to generate specific natural language questions relevant to a user's dataset.
Innovation Solution
A computer-implemented method generates a semantic graph from a user-selected dataset using a generic ontology, identifies relevant central concepts with graph centrality metrics, and assigns relevance scores to paths, outputting natural language questions to guide the user in the data analytics process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If data analysts manually explore data to discover meaningful relationships, then they can understand the dataset thoroughly, but it consumes significant time and slows down the analytics process
Solution Approach 1:
The patent introduces an intermediary system that acts as a mediator between the dataset and the analyst. This system automatically generates natural language questions by mapping data columns to ontology concepts, identifying central concepts using graph centrality metrics, and formulating questions that guide analysts through the data exploration process, thereby reducing manual exploration time while maintaining understanding quality
Solution Approach 2:
The system performs preliminary actions by pre-generating relevant questions based on the dataset's semantic structure before the analyst begins exploration. By pre-computing the semantic graph, identifying central concepts, and formulating guiding questions in advance, the system prepares the analytical pathway ahead of time, eliminating the need for analysts to manually discover data relationships from scratch
2Adaptability or versatility
If current solutions are used for data exploration, then the process follows conventional approaches, but they lack the ability to generate specific natural language questions relevant to the user's dataset
Solution Approach 1:
The patent implements a universal question generation system that can handle multiple types of datasets across different domains through a single platform. The system uses a generic ontology that can map to various data types and domains, and employs universal graph theory algorithms (centrality metrics, path analysis) that work across different dataset structures, enabling one system to serve multiple functions without requiring domain-specific customization
Solution Approach 2:
The system replaces manual mechanical processes of question formulation with automated computational processes. Instead of analysts manually examining data and constructing questions, the system uses automated algorithms to map columns to concepts, compute graph centrality, identify relevant paths, and generate natural language questions programmatically, substituting human cognitive mechanics with computational automation
Data Source
AI summary
Guided exploration of data is provided. A semantic graph corresponding to a dataset is generated using identified relations among columns of the dataset that are identified based on mapping the columns to main concepts in a generic ontology. A subgraph of the semantic graph is formed based on identification of nodes corresponding to relevant central concepts within the semantic graph using graph centrality metrics. A plurality of paths is identified in the subgraph using a bi-directional multiple hop search from the nodes corresponding to the relevant central concepts in the subgraph. A relevance score is assigned to each path in the plurality of paths using a graph-theoretic metric and the graph centrality metrics. A set of natural language questions based on relevant central concepts and concept relations corresponding to nodes in each respective path with an assigned relevance score greater than a minimum threshold score is output to the user.