Document Ontology Generation Using Unsupervised Topic Partitioning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing document management systems rely heavily on manual processes for ontology creation, which are time-consuming, error-prone, and inefficient due to incomplete user understanding, insufficient attribution tools, and large document volumes.
Innovation Solution
An unsupervised machine-learning technique that automatically generates an ontology by partitioning documents based on semantic and structural analysis, using word sequence embedding vectors and clustering algorithms to identify key topics and sub-topics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual processes are used for ontology creation, then user understanding and control are maintained, but the process becomes time-consuming and error-prone
Solution Approach 1:
The system performs self-service by automatically generating ontologies through unsupervised machine learning. The clustering algorithm autonomously analyzes document partitions, computes word sequence embedding vectors, and creates ontology structures without human intervention, eliminating manual errors and time consumption while maintaining reliability through algorithmic consistency
Solution Approach 2:
The patent replaces the mechanical manual process of ontology creation with an automated computational system. Machine learning models and clustering algorithms substitute human analysts, automatically processing documents, extracting topics, and generating ontology structures, thereby resolving the contradiction between manual accuracy and time efficiency
2Measurement precision
If manual attribution tools are used, then some document topics can be identified, but the tools are insufficient for proper and complete attribution
Solution Approach 1:
The system segments documents into multiple partitions and applies clustering algorithms to each partition independently. This segmentation allows the system to identify topics at different granularities and levels of abstraction, improving both measurement precision for specific topics and adaptability for comprehensive coverage across diverse document types
Solution Approach 2:
The patent transforms document attribution from a single-dimensional manual tagging process to a multi-dimensional automated analysis. By computing word sequence embedding vectors and applying clustering in vector space, the system captures semantic relationships and topic structures that manual tools cannot detect, enhancing both precision and completeness of attribution
3Productivity
If automated machine-learning techniques are used, then ontology generation efficiency is improved, but the system complexity increases
Solution Approach 1:
The system segments the complex ontology generation process into distinct manageable stages: document partitioning, word sequence embedding computation, clustering analysis, and ontology structure construction. This segmentation reduces system complexity by breaking down the automated process into modular components while maintaining high productivity through parallel processing of each stage
Data Source
AI summary
Aspects of the present disclosure involve an automated, machine-learning technique for generating a representation of an ontology of a corpus of documents. This unsupervised generation of the ontology of the content of the documents may describe, based on the semantics of the language in the corpus and on the structure and format of the documents in that corpus, potentially key differentiable topics and sub-topics within the documents and the potential relationship between the topics and sub-topics. The unsupervised, or automated, generation of the ontology may provide a foundation of potential topics and sub-topics of a corpus of documents from which a complete ontology for the corpus of documents may be created. This ontology may be both pertinent in defining a structure through which an end user may interpret the data identified from a document or set of documents and/or to inform a machine-learning model to extract document information and classification.