N-gram Machine Learning for Knowledge Graph Generation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Natural Language Processing (NLP) systems that rely on human-curated knowledge bases for semantic parsing are labor-intensive and costly to develop and maintain, making them inefficient for accurately answering natural language questions.
Innovation Solution
A computing system utilizing machine-learned natural language processing models, including an encoder model to generate a knowledge graph from text and a programmer model to produce a program that searches the graph for answers to questions, reducing the need for manual schema development and maintenance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If human-curated knowledge bases with manually defined schema are used for semantic parsing, then accuracy in answering natural language questions is improved, but labor intensity and development cost increase
Solution Approach 1:
The system enables automatic knowledge base construction through machine learning models that autonomously extract entities, relationships, and schema from unstructured text data without requiring manual curation. The encoder-model generates knowledge graphs automatically, and the programmer model creates query programs autonomously, eliminating the need for human experts to manually define schema and populate knowledge bases while maintaining high accuracy in natural language question answering
Solution Approach 2:
The patent replaces the mechanical process of manual knowledge base curation with automated machine learning systems. The encoder model processes text data to generate structured knowledge graphs, and the programmer model automatically generates query programs, substituting human manual work with automated computational processes that scale efficiently without proportional increases in labor cost
2Measurement precision
If human-curated knowledge bases are used for semantic parsing, then answer accuracy is improved, but system complexity and maintenance cost increase
Solution Approach 1:
The system automatically maintains and updates knowledge bases through continuous machine learning processes. The encoder model autonomously extracts and structures information from new text data, updating the knowledge graph without manual intervention. The programmer model automatically adapts to generate appropriate query programs, reducing the complexity of system maintenance while preserving answer accuracy
3Ease of manufacture
If machine-learned models are used to automate knowledge representation, then operational cost is reduced, but the need for automated model training and computation resources increases
Solution Approach 1:
The system performs preliminary action by pre-training the encoder model and programmer model on large datasets before deployment. This upfront computational investment creates optimized models that can then operate efficiently with reduced computation resources during actual natural language question answering tasks, lowering operational costs while maintaining the necessary training infrastructure
Data Source
AI summary
The present disclosure provides systems and methods that perform machine-learned natural language processing. A computing system can include a machine-learned natural language processing model that includes an encoder model trained to receive a natural language text body and output a knowledge graph and a programmer model trained to receive a natural language question and output a program. The computing system can include a computer-readable medium storing instructions that, when executed, cause the processor to perform operations. The operations can include obtaining the natural language text body, inputting the natural language text body into the encoder model, receiving, as an output of the encoder model, the knowledge graph, obtaining the natural language question, inputting the natural language question into the programmer model, receiving the program as an output of the programmer model, and executing the program on the knowledge graph to produce an answer to the natural language question.