Natural Language Query Translation via Knowledge Graph Visualization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing natural language querying systems require detailed knowledge of formal query languages, limiting their applicability, and lack efficient methods for generating ground-truth data needed for machine learning-based translation from natural language to formal queries.
Innovation Solution
A computer-implemented method that provides a knowledge graph as a data model, translates natural language queries into formal data queries, visualizes the queries for user feedback, and stores the queries and feedback as ground-truth pairs, using a visual programming approach and workflow graphs to improve translation accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If formal data query languages are used to query knowledge graphs, then query precision is improved, but ease of operation deteriorates due to requiring detailed knowledge of query language syntax
Solution Approach 1:
The patent introduces a natural language processing intermediary that mediates between the user and the formal query language. Users input natural language queries, which are automatically translated into formal data query language (SPARQL, Cypher, or Gremlin) by the system. This intermediary eliminates the need for users to learn complex query syntax while maintaining precise querying capabilities through the formal language backend.
Solution Approach 2:
The patent replaces the mechanical system of manual query construction with an automated natural language processing system. Instead of requiring users to manually compose queries using formal syntax rules, the system uses machine learning models and natural language understanding to automatically generate and optimize formal queries from casual user input, substituting automated intelligence for manual mechanical query building.
2Ease of operation
If machine learning techniques are used to translate natural language queries into formal data queries, then ease of operation is improved, but reliability deteriorates due to lack of ground-truth training data
Solution Approach 1:
The patent performs preliminary action by pre-collecting and curating ground-truth query pairs before deploying the machine learning model. The system gathers examples of natural language queries paired with their correct formal query translations, stores them in a training dataset, and uses this pre-prepared data to train the model. This preliminary data preparation ensures the model has high-quality training material before actual deployment, improving reliability from the outset.
Solution Approach 2:
The patent implements a self-service mechanism where the system automatically generates additional ground-truth data by executing formal queries against the knowledge graph and observing results. The system uses these execution outcomes to refine and validate its translations, creating a self-improving loop where the model learns from its own performance feedback without requiring constant external annotation, thereby improving reliability over time.
3Reliability
If ground-truth data is manually created for training machine learning models, then reliability is improved, but productivity deteriorates due to time-consuming manual annotation
Solution Approach 1:
The system employs self-service by automatically generating ground-truth training data through formal query execution on the knowledge graph. Instead of relying on manual annotation, the system executes generated queries, captures the results, and uses these actual execution outcomes as ground truth for training and validation. This automated self-generation process dramatically improves productivity while maintaining high reliability since the ground truth comes from actual system performance rather than subjective human annotation.
Solution Approach 2:
The patent applies parameter changes by transforming the state of query data from unverified translations to executed and validated results. By changing the parameter of query verification from manual review to automated execution, the system rapidly generates large volumes of reliable ground-truth data. This parameter change from static manual annotation to dynamic automated validation exponentially increases productivity while ensuring the reliability of training data through actual system behavior observation.
Data Source
AI summary
A computer-implemented method for generating ground-truth for natural language querying may include providing a knowledge graph as data model, receiving a natural language query from a user and translating the natural language query into a formal data query. The method can also include visualizing the formal data query to the user and receiving a feedback response from the user. The feedback response can include a verified and/or edited formal data query. The method can also include storing the natural language query and the corresponding feedback response as ground-truth pair. Corresponding system and a related computer program product may be provided.